id,name,unit,description,createdAt,updatedAt,code,coverage,timespan,datasetId,sourceId,shortUnit,display,columnOrder,originalMetadata,grapherConfigAdmin,shortName,catalogPath,dimensions,schemaVersion,processingLevel,processingLog,titlePublic,titleVariant,attributionShort,attribution,descriptionShort,descriptionFromProducer,descriptionKey,descriptionProcessing,licenses,license,grapherConfigETL,type,sort,dataChecksum,metadataChecksum
954464,Patent applications granted per 1 million people - Field: Transportation,granted applications per 1 million people,,2024-07-25 13:04:21,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""granted applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_per_mil__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_per_mil__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Patent applications granted per 1 million people"", ""originalShortName"": ""num_patent_granted_per_mil""}",2,major,,,,,,Yearly total number of patents granted per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],0505b489a2c963e1387f15b02d3829c1,32a5329034b8b7d37e5c5b94eacc9151
954463,Patent applications per 1 million people - Field: Transportation,applications per 1 million people,,2024-07-25 13:04:21,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_per_mil__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_per_mil__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Patent applications per 1 million people"", ""originalShortName"": ""num_patent_applications_per_mil""}",2,major,,,,,,Yearly total number of patent applications per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],5a4c577480a5262f01abcedd1c323c17,c2c215d7303b633af1d33fc2f4aaf5dc
954462,Total patent applications granted - Field: Transportation,granted applications,,2024-07-25 13:04:21,2024-07-30 06:18:30,,,2023-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_summary__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_summary__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Total patent applications granted"", ""originalShortName"": ""num_patent_granted_summary""}",2,,,,,,,Total number of patents granted between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],a144a2cf15a79ea6c4b31080e01f36c4,71a982828d7c8d5915657d30152169c9
954461,Total patent applications - Field: Transportation,applications,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2023-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_summary__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_summary__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Total patent applications"", ""originalShortName"": ""num_patent_applications_summary""}",2,,,,,,,Total number of patent applications between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],00b9239206c702da39b19f1b88c7f516,3ff5eb442c377d3cb1d5e269211204ea
954460,Patent applications granted - Field: Transportation,granted applications,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Patent applications granted"", ""originalShortName"": ""num_patent_granted""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],9fe4008b113c6ff1269c806ceb7d85ce,342ddaeb9676147318150a9653463aa6
954459,Patent applications - Field: Transportation,applications,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Patent applications"", ""originalShortName"": ""num_patent_applications""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],95d7216fcae200efd243a9e6443b52bc,142c951b1cd3b2669852a2dae359e4df
954458,Total estimated investment - Field: Transportation,constant 2021 US$,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],9e735203bd41702b0383aa77c8811a44,ed281cd8220187ab2f8e8f3d56861267
954457,Total disclosed investment - Field: Transportation,constant 2021 US$,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],a52d7b6ba2c83bfb224e5574b2c16337,087fb13cc29bf8de772fc31664a68244
954456,Estimated investment - Field: Transportation,constant 2021 US$,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],a3aa74278f99e069f2583e7968d202c4,b82f2253f25dba31bd4afcbb0270a83b
954455,Disclosed investment - Field: Transportation,constant 2021 US$,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_transportation,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_transportation,"{""filters"": [{""name"": ""field"", ""value"": ""Transportation""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],72aa172fd9daf77883122e7aca116a56,75e6c395f3676dce270cb51a6a721526
954454,"Citations per article, 2013-2023 - Field: Theoretical computer science",citations per article,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2023-2023,6633,,,"{""unit"": ""citations per article"", ""numDecimalPlaces"": 0}",0,,,citations_per_article__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#citations_per_article__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Citations per article, 2013-2023"", ""originalShortName"": ""citations_per_article""}",2,,,,,,,"The total number of citations per article in the AI field between 2013-2023. Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],"Countries that have published fewer than 1,000 articles on AI between 2010-2023 are excluded to prevent distortion in the ""citations per publication"" metric, which can occur when dividing by a small sample size.",,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",float,[],eabef5a01df58ba89c195d0e57d0f0c0,f0f04836ef64045f0f45f8f5ab88231e
954453,Number of articles per 1 million people - Field: Theoretical computer science,articles per 1 million people,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""articles per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_articles_per_mil__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_per_mil__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Number of articles per 1 million people"", ""originalShortName"": ""num_articles_per_mil""}",2,major,,,,,,"Yearly number of articles published in AI fields per 1 million people. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""Data for 2022-2023 is incomplete.""}",float,[],6d7883f1cab8371964f25f3cab820d85,c5ef829c7f2184ed1717d4ebf78396d4
954452,Total number of citations - Field: Theoretical computer science,citations,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2023-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations_summary__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations_summary__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Total number of citations"", ""originalShortName"": ""num_citations_summary""}",2,,,,,,,"Total number of citations in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],c5436d2ce35fddab1d9b249824114b58,5d9e672bb923c7b986bd65603dd7cd8d
954451,Total number of articles - Field: Theoretical computer science,articles,,2024-07-25 13:04:20,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles_summary__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_summary__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Total number of articles"", ""originalShortName"": ""num_articles_summary""}",2,,,,,,,"Total number of articles in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],fd8d43c1f7575ffc56d79212305f86d9,6c53d27f4a6aa66ec4d005334b9a2a47
954450,Number of citations - Field: Theoretical computer science,citations,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Number of citations"", ""originalShortName"": ""num_citations""}",2,,,,,,,"Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],2dc9f8dd8476d424f3bb7619b870861b,5b5dd69ee0ed55351226a77d6362dd4d
954449,Number of articles - Field: Theoretical computer science,articles,,2024-07-25 13:04:20,2024-07-30 06:18:30,,,2013-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles__field_theoretical_computer_science,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles__field_theoretical_computer_science,"{""filters"": [{""name"": ""field"", ""value"": ""Theoretical computer science""}], ""originalName"": ""Number of articles"", ""originalShortName"": ""num_articles""}",2,,,,,,,"English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],50e86eb78e7317ca1c35756e2bfac2b3,ce1324a600ce9a39582c11de4ffb64c2
954448,Patent applications granted per 1 million people - Field: Telecommunications,granted applications per 1 million people,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""granted applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_per_mil__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_per_mil__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Patent applications granted per 1 million people"", ""originalShortName"": ""num_patent_granted_per_mil""}",2,major,,,,,,Yearly total number of patents granted per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],4d07178ab48eb43068a5beeaee67c4ca,b63b218759ee1abf0e3545af9c6142cf
954447,Patent applications per 1 million people - Field: Telecommunications,applications per 1 million people,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_per_mil__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_per_mil__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Patent applications per 1 million people"", ""originalShortName"": ""num_patent_applications_per_mil""}",2,major,,,,,,Yearly total number of patent applications per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],0c8ee9ef0d44b63613266addfaeb215a,d69319711f730bafe9b2c3f0e4e223fb
954446,Total patent applications granted - Field: Telecommunications,granted applications,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_summary__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_summary__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Total patent applications granted"", ""originalShortName"": ""num_patent_granted_summary""}",2,,,,,,,Total number of patents granted between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],9716ed9eeb89d57923d6ec635c822603,59a5193323f10185b8b984bba5745122
954445,Patent applications granted - Field: Telecommunications,granted applications,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Patent applications granted"", ""originalShortName"": ""num_patent_granted""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],9f795b3cb50243ac33fbf21714003f6e,5e4912ec947d415528584be0124ac7a1
954444,Patent applications - Field: Telecommunications,applications,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Patent applications"", ""originalShortName"": ""num_patent_applications""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],97d6fc324cf17081a201c9a7c1a99b03,3c8c913b54341c6fdcb57f9646b212dd
954443,Total patent applications - Field: Telecommunications,applications,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_summary__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_summary__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Total patent applications"", ""originalShortName"": ""num_patent_applications_summary""}",2,,,,,,,Total number of patent applications between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],a68345cdc3a0e221bea76f0ce5668545,d177f32258f2b067be12d98caf3dad10
954442,Total estimated investment - Field: Telecommunications,constant 2021 US$,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],2b25b8d41b45f8ab09c903a758bf04a4,5b7e4592d6499f54ec48fe9f17d31b00
954441,Total disclosed investment - Field: Telecommunications,constant 2021 US$,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],7f7f8a1e5cb7f7f4e0bb0aeda59d63ce,9252124494d0cdbe2195873d33b140b7
954440,Estimated investment - Field: Telecommunications,constant 2021 US$,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],c9ecddae93c5183104a1f15ea7cb5be0,82db544ffae960f72069f30b5e954294
954439,Disclosed investment - Field: Telecommunications,constant 2021 US$,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_telecommunications,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_telecommunications,"{""filters"": [{""name"": ""field"", ""value"": ""Telecommunications""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],85efa708e44292f136a1a13f3c626873,d347c69ee3248255cb141047996aa1f8
954438,Total estimated investment - Field: Sports,constant 2021 US$,,2024-07-25 13:04:19,2024-07-30 06:18:29,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_sports,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_sports,"{""filters"": [{""name"": ""field"", ""value"": ""Sports""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],8a3e29bdd3d4755643eb493d4de3585f,3fce09325593ad61fa49764a2dcd057e
954437,Total disclosed investment - Field: Sports,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_sports,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_sports,"{""filters"": [{""name"": ""field"", ""value"": ""Sports""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],9a937b6d59e94c81935f00d1ad6e7841,ccb38d69683fe0df3ad5ba45c96b22c9
954436,Estimated investment - Field: Sports,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_sports,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_sports,"{""filters"": [{""name"": ""field"", ""value"": ""Sports""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],8f48dacb6e94a048b92fd959c8b666c8,7223397c5cf8aab44a20b176db468b52
954435,Disclosed investment - Field: Sports,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_sports,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_sports,"{""filters"": [{""name"": ""field"", ""value"": ""Sports""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],86a9d423f827f7f6eca625b70cd7203a,c74b127fc3cef8919d81678c041b0104
954434,"Citations per article, 2013-2023 - Field: Speech recognition",citations per article,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""citations per article"", ""numDecimalPlaces"": 0}",0,,,citations_per_article__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#citations_per_article__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Citations per article, 2013-2023"", ""originalShortName"": ""citations_per_article""}",2,,,,,,,"The total number of citations per article in the AI field between 2013-2023. Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],"Countries that have published fewer than 1,000 articles on AI between 2010-2023 are excluded to prevent distortion in the ""citations per publication"" metric, which can occur when dividing by a small sample size.",,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",float,[],705d0b1fbba9019216a874d1ad69330c,699d2f7e5936ef6d1b329ffff6f6baa1
954433,Number of articles per 1 million people - Field: Speech recognition,articles per 1 million people,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""articles per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_articles_per_mil__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_per_mil__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Number of articles per 1 million people"", ""originalShortName"": ""num_articles_per_mil""}",2,major,,,,,,"Yearly number of articles published in AI fields per 1 million people. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""Data for 2022-2023 is incomplete.""}",float,[],23db7706138eb56ffa6513166b980a55,520750a18f21a144f14c000101242e19
954432,Total number of citations - Field: Speech recognition,citations,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations_summary__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations_summary__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Total number of citations"", ""originalShortName"": ""num_citations_summary""}",2,,,,,,,"Total number of citations in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],6a31064d6bb83524c4d9adaad2c1c17b,8083583ae8233f0ec6c824de1701433d
954431,Total number of articles - Field: Speech recognition,articles,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2023-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles_summary__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_summary__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Total number of articles"", ""originalShortName"": ""num_articles_summary""}",2,,,,,,,"Total number of articles in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],569e060f31ab028f9dbadae936e3731e,06b85a83e1794d68d47ff5fae9272cbf
954430,Number of citations - Field: Speech recognition,citations,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Number of citations"", ""originalShortName"": ""num_citations""}",2,,,,,,,"Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],617b41921b679bb505889df75694e6e3,3dcfb2513ef7b8da677e9f7655a2e55c
954429,Number of articles - Field: Speech recognition,articles,,2024-07-25 13:04:18,2024-07-30 06:18:29,,,2013-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Number of articles"", ""originalShortName"": ""num_articles""}",2,,,,,,,"English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],3ca731269521bf116b3514b1b547558e,4dc07c989e8e42f5c4eb61bb1d4c8524
954428,Total estimated investment - Field: Speech recognition,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],11d2f545a4e330c341321620f7c7edf4,e14facbd6995dcc568d03c924f8133e3
954427,Total disclosed investment - Field: Speech recognition,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],904e79f1e3e1fcad11ce9d283a5bce96,c141d694547d9563e289609c5d452dfc
954426,Estimated investment - Field: Speech recognition,constant 2021 US$,,2024-07-25 13:04:18,2024-07-30 06:18:28,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],bde18b150fc03ff7a7fb318ea887e1b4,2e14aa30becd7b561ec6ce0640662b37
954425,Disclosed investment - Field: Speech recognition,constant 2021 US$,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_speech_recognition,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_speech_recognition,"{""filters"": [{""name"": ""field"", ""value"": ""Speech recognition""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],8c05245b42d15bdfd6880249ff9dc993,4c608b8ab67c51c5205bb522f9b4ebe4
954424,Patent applications granted per 1 million people - Field: Speech processing,granted applications per 1 million people,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""granted applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_per_mil__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_per_mil__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Patent applications granted per 1 million people"", ""originalShortName"": ""num_patent_granted_per_mil""}",2,major,,,,,,Yearly total number of patents granted per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],200149a4a5c2ab454a1a7712030f30ea,f1ee09ef20b2c3f9c82cdc3f3496cc15
954423,Patent applications per 1 million people - Field: Speech processing,applications per 1 million people,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_per_mil__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_per_mil__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Patent applications per 1 million people"", ""originalShortName"": ""num_patent_applications_per_mil""}",2,major,,,,,,Yearly total number of patent applications per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],3fa89a2da86f7b5f7ff6de578678f4c5,a3ad02878c5d085d600e1ac256ec03af
954422,Total patent applications granted - Field: Speech processing,granted applications,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_summary__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_summary__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Total patent applications granted"", ""originalShortName"": ""num_patent_granted_summary""}",2,,,,,,,Total number of patents granted between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],762b47035f110be6725ab6e9e7d00ff2,f8614c8599c4bb98a5cd038293acb4ff
954421,Total patent applications - Field: Speech processing,applications,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_summary__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_summary__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Total patent applications"", ""originalShortName"": ""num_patent_applications_summary""}",2,,,,,,,Total number of patent applications between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],4b95de07fc3008697d2e6fe9cc22799a,ecca9f3bbe273e8673f02da42374e043
954420,Patent applications granted - Field: Speech processing,granted applications,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Patent applications granted"", ""originalShortName"": ""num_patent_granted""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],b15fae820c63fdee68662f72f23ca5ed,3ebd3ba8641c2ac5d906eaca7b736974
954419,Patent applications - Field: Speech processing,applications,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications__field_speech_processing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications__field_speech_processing,"{""filters"": [{""name"": ""field"", ""value"": ""Speech processing""}], ""originalName"": ""Patent applications"", ""originalShortName"": ""num_patent_applications""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],ff901cd6844f7924d60b16b00b4e6570,66aba614b187bbf5998f3035ad2c5abd
954418,Total estimated investment - Field: Software,constant 2021 US$,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_software,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_software,"{""filters"": [{""name"": ""field"", ""value"": ""Software""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],1fe93210578d64cc4c626fee53cb693a,6699d4b2ab0f6fe880804cdf77a9e537
954417,Total disclosed investment - Field: Software,constant 2021 US$,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_software,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_software,"{""filters"": [{""name"": ""field"", ""value"": ""Software""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],d935b3c93d95d5a71da4e1be2787e309,8c2c616d028432fbd69cf25e2f0e5d57
954416,Estimated investment - Field: Software,constant 2021 US$,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_software,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_software,"{""filters"": [{""name"": ""field"", ""value"": ""Software""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],0292fdf12985eb0d522b6eae98fff46c,93d5fbaf46d895ae710f2efee6458861
954415,Disclosed investment - Field: Software,constant 2021 US$,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_software,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_software,"{""filters"": [{""name"": ""field"", ""value"": ""Software""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],20e3d335eeb7492d1de5d48ec08827e2,6d6bf5bf64045e68973b2635892888f2
954414,"Citations per article, 2013-2023 - Field: Simulation",citations per article,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,,"{""unit"": ""citations per article"", ""numDecimalPlaces"": 0}",0,,,citations_per_article__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#citations_per_article__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Citations per article, 2013-2023"", ""originalShortName"": ""citations_per_article""}",2,,,,,,,"The total number of citations per article in the AI field between 2013-2023. Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],"Countries that have published fewer than 1,000 articles on AI between 2010-2023 are excluded to prevent distortion in the ""citations per publication"" metric, which can occur when dividing by a small sample size.",,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",float,[],b5e0fb1666d43d99b3b9f044b7f4c6f7,97260da48a93cf3918700142f8b8bcfe
954413,Number of articles per 1 million people - Field: Simulation,articles per 1 million people,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""articles per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_articles_per_mil__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_per_mil__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Number of articles per 1 million people"", ""originalShortName"": ""num_articles_per_mil""}",2,major,,,,,,"Yearly number of articles published in AI fields per 1 million people. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""Data for 2022-2023 is incomplete.""}",float,[],1a1beceaa5ec49d360b3c6062bcf1156,e5c9e91b9fbd2437b855dbeef96f9353
954412,Total number of citations - Field: Simulation,citations,,2024-07-25 13:04:17,2024-07-30 06:18:28,,,2023-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations_summary__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations_summary__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Total number of citations"", ""originalShortName"": ""num_citations_summary""}",2,,,,,,,"Total number of citations in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],0d7e77d939c7e4f964a2b56f2a87809a,c419209fde5ceaeab4961de41837ad20
954411,Total number of articles - Field: Simulation,articles,,2024-07-25 13:04:16,2024-07-30 06:18:28,,,2023-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles_summary__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_summary__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Total number of articles"", ""originalShortName"": ""num_articles_summary""}",2,,,,,,,"Total number of articles in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],58863155b795bd8717df56f53f3c73de,7ca7f4c0de50df835256a7244e4e0fb3
954410,Number of citations - Field: Simulation,citations,,2024-07-25 13:04:16,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Number of citations"", ""originalShortName"": ""num_citations""}",2,,,,,,,"Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],9d21bbc53d561f1a58897d9c90580ee7,ca9448ad4db53e4526a4c0619d306e4b
954409,Number of articles - Field: Simulation,articles,,2024-07-25 13:04:16,2024-07-30 06:18:28,,,2013-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Number of articles"", ""originalShortName"": ""num_articles""}",2,,,,,,,"English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],7907f209c472e62d3b4ca4d8a4d539a9,b74ada07c4187d6e821a185f17496d69
954408,Total estimated investment - Field: Simulation,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],a2d5ee8af70fa83d41975d812ffa4b44,6205bbdd3c730de0233468a59e3ad120
954407,Total disclosed investment - Field: Simulation,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:28,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],e796e60c87c9120ca6af1aea8e4dfe6d,235bfce78cf3186556ab038594825ffd
954406,Estimated investment - Field: Simulation,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],4b541711c777248734c16473f0e5147d,09f4b56cf78174eb801b7522db9a8fc5
954405,Disclosed investment - Field: Simulation,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_simulation,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_simulation,"{""filters"": [{""name"": ""field"", ""value"": ""Simulation""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],088c5a4f3c0410d1231cb5bcb6d7e611,63f40a97f67a79d74674045215ae0bfc
954404,Total estimated investment - Field: Semiconductor,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_semiconductor,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_semiconductor,"{""filters"": [{""name"": ""field"", ""value"": ""Semiconductor""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],ca8d594839712ca5e1bafa6e167c35c2,13348f662501c20b9bea744c4582ec9d
954403,Total disclosed investment - Field: Semiconductor,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_semiconductor,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_semiconductor,"{""filters"": [{""name"": ""field"", ""value"": ""Semiconductor""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],dca3adbca21489f55966fc6c69a3a07d,ae717cc42238938ae3cfc01734c6d55c
954402,Estimated investment - Field: Semiconductor,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_semiconductor,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_semiconductor,"{""filters"": [{""name"": ""field"", ""value"": ""Semiconductor""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],35bc310049d36c19a67e6883f27e4333,bd76529d4261e5201b9556de974ab4fa
954401,Disclosed investment - Field: Semiconductor,constant 2021 US$,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_semiconductor,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_semiconductor,"{""filters"": [{""name"": ""field"", ""value"": ""Semiconductor""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],bf9a16d284ea790bf0b4ea0245e8f0c7,3b4ec2f26315b0e8e51dfbfc3f929527
954400,Patent applications granted per 1 million people - Field: Security,granted applications per 1 million people,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""granted applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_per_mil__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_per_mil__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Patent applications granted per 1 million people"", ""originalShortName"": ""num_patent_granted_per_mil""}",2,major,,,,,,Yearly total number of patents granted per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],2a77dce78f0edd572cdfc38c17422049,63efb076a19999cee3e27035f81ad290
954399,Patent applications per 1 million people - Field: Security,applications per 1 million people,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_per_mil__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_per_mil__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Patent applications per 1 million people"", ""originalShortName"": ""num_patent_applications_per_mil""}",2,major,,,,,,Yearly total number of patent applications per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],71473c8e7b84657bd0df6cb3b0769f31,acae53d05a7e9e8309035e420f25ed86
954398,Total patent applications granted - Field: Security,granted applications,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2023-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_summary__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_summary__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Total patent applications granted"", ""originalShortName"": ""num_patent_granted_summary""}",2,,,,,,,Total number of patents granted between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],e82f80a7f80c16fb554606f714b2be7e,c975ff987f59805af59b79dee3712e82
954397,Total patent applications - Field: Security,applications,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2023-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_summary__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_summary__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Total patent applications"", ""originalShortName"": ""num_patent_applications_summary""}",2,,,,,,,Total number of patent applications between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],2b461c3f487548cd948d865fc4d55611,3615f5e419a94d3b01a61d8e678a249a
954396,Patent applications granted - Field: Security,granted applications,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Patent applications granted"", ""originalShortName"": ""num_patent_granted""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],f041e062c7078ffb14676e1f8ea1e438,b1ee9386654f3c6cef1a3a4e5f139df6
954395,Patent applications - Field: Security,applications,,2024-07-25 13:04:16,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications__field_security,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications__field_security,"{""filters"": [{""name"": ""field"", ""value"": ""Security""}], ""originalName"": ""Patent applications"", ""originalShortName"": ""num_patent_applications""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],64070eac17644ad4da80a22bc277e1c6,b3c5c137d0b1bffe500c3af3600252b3
954394,"Total estimated investment - Field: Sales, retail, commerce, marketing",constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_sales__retail__commerce__marketing,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_sales__retail__commerce__marketing,"{""filters"": [{""name"": ""field"", ""value"": ""Sales, retail, commerce, marketing""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],beca30658e64848177d631a0b2e1721c,a81709d41e12944d2379ec836e4864f1
954393,"Total disclosed investment - Field: Sales, retail, commerce, marketing",constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_sales__retail__commerce__marketing,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_sales__retail__commerce__marketing,"{""filters"": [{""name"": ""field"", ""value"": ""Sales, retail, commerce, marketing""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],38ef8df4434286736e614e92009c73bd,680d8161ce2e368cdc76ff5f4338c20b
954392,"Estimated investment - Field: Sales, retail, commerce, marketing",constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_sales__retail__commerce__marketing,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_sales__retail__commerce__marketing,"{""filters"": [{""name"": ""field"", ""value"": ""Sales, retail, commerce, marketing""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],f5b3a013994317368bbc4c48ea83a312,bfc5d9d866efcb60e334fb6a12b5b45d
954391,"Disclosed investment - Field: Sales, retail, commerce, marketing",constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_sales__retail__commerce__marketing,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_sales__retail__commerce__marketing,"{""filters"": [{""name"": ""field"", ""value"": ""Sales, retail, commerce, marketing""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],f48bfc665eab7d92e873af1556fe9039,adbcd8665c407cb5c482dba5be8204e0
954390,Total estimated investment - Field: Robotics,constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_robotics,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_robotics,"{""filters"": [{""name"": ""field"", ""value"": ""Robotics""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],e4b1d263f307ded2131580bec6c9a458,94a6fba130c870b0021bc8362e439836
954389,Total disclosed investment - Field: Robotics,constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_robotics,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_robotics,"{""filters"": [{""name"": ""field"", ""value"": ""Robotics""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],cc3e45d48cb6d4918e99a83da26fb79c,5b014152a5961dcc70a8b579d8d4eb71
954388,Estimated investment - Field: Robotics,constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_robotics,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_robotics,"{""filters"": [{""name"": ""field"", ""value"": ""Robotics""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],57f4248bf67d1aca3740cddbff9415fb,91c0322d18bec24eb64fcc62969d3aa6
954387,Disclosed investment - Field: Robotics,constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_robotics,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_robotics,"{""filters"": [{""name"": ""field"", ""value"": ""Robotics""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],96212bb526d0952d72deb6cbf0f52533,4a4c9f32f4e16e58e42d5361ffe23fff
954386,"Citations per article, 2013-2023 - Field: Real-time computing",citations per article,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2023-2023,6633,,,"{""unit"": ""citations per article"", ""numDecimalPlaces"": 0}",0,,,citations_per_article__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#citations_per_article__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Citations per article, 2013-2023"", ""originalShortName"": ""citations_per_article""}",2,,,,,,,"The total number of citations per article in the AI field between 2013-2023. Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],"Countries that have published fewer than 1,000 articles on AI between 2010-2023 are excluded to prevent distortion in the ""citations per publication"" metric, which can occur when dividing by a small sample size.",,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",float,[],27c0cc086d317a170f00512abe5065c9,b38682f0f8a4fcb332eab94326cd04bb
954385,Number of articles per 1 million people - Field: Real-time computing,articles per 1 million people,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""articles per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_articles_per_mil__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_per_mil__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Number of articles per 1 million people"", ""originalShortName"": ""num_articles_per_mil""}",2,major,,,,,,"Yearly number of articles published in AI fields per 1 million people. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""Data for 2022-2023 is incomplete.""}",float,[],c5a1bd063ff142bd7721326e7c8b4954,3f9d56957482b476116bf129981e9a68
954384,Total number of citations - Field: Real-time computing,citations,,2024-07-25 13:04:15,2024-07-30 06:18:26,,,2023-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations_summary__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations_summary__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Total number of citations"", ""originalShortName"": ""num_citations_summary""}",2,,,,,,,"Total number of citations in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],ce0c89dc04556a3a3d41bfdfa69cca0d,8ec192487cfeb8efccb2b0dae9367e6c
954383,Total number of articles - Field: Real-time computing,articles,,2024-07-25 13:04:15,2024-07-30 06:18:26,,,2023-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles_summary__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles_summary__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Total number of articles"", ""originalShortName"": ""num_articles_summary""}",2,,,,,,,"Total number of articles in the AI field between 2013-2023. English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],69505d47b262b7b552730300536bff3c,20d11dccfc2cc6434dc9ce2b547ac5d5
954382,Number of citations - Field: Real-time computing,citations,,2024-07-25 13:04:15,2024-07-30 06:18:27,,,2013-2023,6633,,,"{""unit"": ""citations"", ""numDecimalPlaces"": 0}",0,,,num_citations__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_citations__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Number of citations"", ""originalShortName"": ""num_citations""}",2,,,,,,,"Citations take time to accumulate so recent years may have fewer citations, and the measure for these years may not completely reflective of their eventual total.",,[],,,,"{""note"": ""The citation counts for papers published in a specific year are subject to a time lag, as papers often accrue most citations in subsequent years. Data for 2022-2023 is therefore incomplete.""}",int,[],0c1b56766b8a97b373ea680bd00a3614,997f19b8747494bb89ad1fd7b615c9ca
954381,Number of articles - Field: Real-time computing,articles,,2024-07-25 13:04:15,2024-07-30 06:18:26,,,2013-2023,6633,,,"{""unit"": ""articles"", ""numDecimalPlaces"": 0}",0,,,num_articles__field_real_time_computing,grapher/artificial_intelligence/2024-07-16/cset/cset#num_articles__field_real_time_computing,"{""filters"": [{""name"": ""field"", ""value"": ""Real-time computing""}], ""originalName"": ""Number of articles"", ""originalShortName"": ""num_articles""}",2,,,,,,,"English- and Chinese-language scholarly publications related to the development and application of AI. This includes journal articles, conference papers, repository publications (such as arXiv), books, and theses.",,[],,,,"{""note"": ""Data for 2022-2023 is incomplete.""}",int,[],abf71d03102336ebe932b8733f1914a2,8e654a2ca35b1d7c0efa8162c7640b72
954380,Total estimated investment - Field: Real estate,constant 2021 US$,,2024-07-25 13:04:15,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_real_estate,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_real_estate,"{""filters"": [{""name"": ""field"", ""value"": ""Real estate""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],14b281c7c18415730dba605c95b48d10,32b5dc92f6179fec96377f6cc7e80c0f
954379,Total disclosed investment - Field: Real estate,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_real_estate,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_real_estate,"{""filters"": [{""name"": ""field"", ""value"": ""Real estate""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],9507df5a7db6370ddf175619b4e28fd0,5a78cca3cced1543365a03539f896ec6
954378,Estimated investment - Field: Real estate,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_real_estate,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_real_estate,"{""filters"": [{""name"": ""field"", ""value"": ""Real estate""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],cb6b86a1b8b87880a62abd700422d9e2,c4d5eeb03d9d595988dd068fe3badbb0
954377,Disclosed investment - Field: Real estate,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_real_estate,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_real_estate,"{""filters"": [{""name"": ""field"", ""value"": ""Real estate""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],a26af7e277221aa140bca315c38d07eb,387c6e03338d020e21a2430e38d79a7f
954376,Total estimated investment - Field: Professional services,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_professional_services,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_professional_services,"{""filters"": [{""name"": ""field"", ""value"": ""Professional services""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],fcb54afd687bfab67a6363cceff07c22,70756f6a1ce6600c0841d8f1d887b71b
954375,Total disclosed investment - Field: Professional services,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_professional_services,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_professional_services,"{""filters"": [{""name"": ""field"", ""value"": ""Professional services""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],de7b7c213462e67e2992d1e71c7abce1,7d223c46de148d1caebc2270e2c39b84
954374,Estimated investment - Field: Professional services,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,investment_estimated__field_professional_services,grapher/artificial_intelligence/2024-07-16/cset/cset#investment_estimated__field_professional_services,"{""filters"": [{""name"": ""field"", ""value"": ""Professional services""}], ""originalName"": ""Estimated investment"", ""originalShortName"": ""investment_estimated""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],8391b144d827ccca7a0193fddca26c99,b8e41e3c8866f29d30f9c9b41cbec4fa
954373,Disclosed investment - Field: Professional services,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment__field_professional_services,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment__field_professional_services,"{""filters"": [{""name"": ""field"", ""value"": ""Professional services""}], ""originalName"": ""Disclosed investment"", ""originalShortName"": ""disclosed_investment""}",2,,,,,,,"Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],24ca3a50ad319a5531182ac539ef1ece,552c65390a93487ea6e7729bc9c68a73
954372,Patent applications granted per 1 million people - Field: Probabilistic reasoning,granted applications per 1 million people,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,,"{""unit"": ""granted applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_per_mil__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_per_mil__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Patent applications granted per 1 million people"", ""originalShortName"": ""num_patent_granted_per_mil""}",2,major,,,,,,Yearly total number of patents granted per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],46b0511361798c126f91e095fd7a0081,c8b38217c4de88f3f3dec761c97a82d1
954371,Patent applications per 1 million people - Field: Probabilistic reasoning,applications per 1 million people,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,,"{""unit"": ""applications per 1 million people"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_per_mil__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_per_mil__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Patent applications per 1 million people"", ""originalShortName"": ""num_patent_applications_per_mil""}",2,major,,,,,,Yearly total number of patent applications per 1 million people. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,"[{""url"": ""https://docs.google.com/document/d/1-RmthhS2EPMK_HIpnPctcXpB0n7ADSWnXa5Hb3PxNq4/edit?usp=sharing"", ""name"": ""Creative Commons BY 4.0""}]",,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",float,[],4557f5f180eb0559c981a44148a91190,2d297dab146296f735287bf34f49beea
954370,Total patent applications granted - Field: Probabilistic reasoning,granted applications,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted_summary__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted_summary__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Total patent applications granted"", ""originalShortName"": ""num_patent_granted_summary""}",2,,,,,,,Total number of patents granted between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],cae3b7d4a69cee48b2c893fbf97b10e1,05ebfbfedabba25af2108cdf8b9958b2
954369,Total patent applications - Field: Probabilistic reasoning,applications,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications_summary__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications_summary__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Total patent applications"", ""originalShortName"": ""num_patent_applications_summary""}",2,,,,,,,Total number of patent applications between 2013-2023. Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],8c26acee432b6251ffe67624f3b5893a,69aaaca4341e6ec04bf62bb06ed0819e
954368,Patent applications granted - Field: Probabilistic reasoning,granted applications,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,,"{""unit"": ""granted applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_granted__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_granted__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Patent applications granted"", ""originalShortName"": ""num_patent_granted""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],ae39e2a9a849dafcbcad0caa28cd9aec,10b5a1d188b31e9b733486bc73d7ffc9
954367,Patent applications - Field: Probabilistic reasoning,applications,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2013-2023,6633,,,"{""unit"": ""applications"", ""numDecimalPlaces"": 0}",0,,,num_patent_applications__field_probabilistic_reasoning,grapher/artificial_intelligence/2024-07-16/cset/cset#num_patent_applications__field_probabilistic_reasoning,"{""filters"": [{""name"": ""field"", ""value"": ""Probabilistic reasoning""}], ""originalName"": ""Patent applications"", ""originalShortName"": ""num_patent_applications""}",2,,,,,,,Patents related to artificial intelligence first submitted in the selected country's patent office. Subsequent granting of that patent could be by any country's patent office.,,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Bermuda, Armenia, Belarus, Georgia, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Iceland, Albania, Andorra, Bosnia and Herzegovina, Malta, Montenegro, San Marino, North Macedonia, Liechtenstein, Monaco, Vatican City, Guernsey, Afghanistan, Kyrgyzstan, Bahrain, Laos, Bangladesh, Lebanon, Bhutan, Maldives, Cambodia, Syria, Tajikistan, Cyprus, Mongolia, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Pakistan, Palestine, Iraq, United Arab Emirates, Uzbekistan, Kazakhstan, Qatar, Vietnam, Yemen, Kuwait, Algeria, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Nigeria, Botswana, Gambia, Rwanda, Burkina Faso, Ghana, São Tomé and Príncipe, Burundi, Guinea, Senegal, Guinea-Bissau, Seychelles, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Madagascar, Sudan, Côte d'Ivoire, Malawi, Togo, Mali, Djibouti, Mauritania, Uganda, Egypt, Mauritius, Tanzania, Zambia, Eritrea, Mozambique, Zimbabwe, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Dominican Republic, Bahamas, Ecuador, Paraguay, Barbados, Saint Vincent and the Grenadines, El Salvador, Belize, Grenada, Saint Kitts and Nevis, Guatemala, Guyana, Haiti, Honduras, Trinidad and Tobago, Jamaica, Venezuela, Puerto Rico, Cayman Islands (the), Turks and Caicos Islands, Saint Lucia, and Dominica.""]",,,,"{""note"": ""According to calculations by CSET, the median time for a patent to be granted is 826 days from its initial filing date, while the average time is 860 days.  Data for 2021-2023 is incomplete.""}",int,[],a713fd6104e3d0427d63159d65a50df4,6f325c2fb100473382ca26c034958c71
954366,Total estimated investment - Field: Privacy and security,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,estimated_investment_summary__field_privacy_and_security,grapher/artificial_intelligence/2024-07-16/cset/cset#estimated_investment_summary__field_privacy_and_security,"{""filters"": [{""name"": ""field"", ""value"": ""Privacy and security""}], ""originalName"": ""Total estimated investment"", ""originalShortName"": ""estimated_investment_summary""}",2,,,,,,,"Total estimated investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],3a05dcf79f1844f5ce4477380f082609,2aaba7e8276550fbbe43e5d8ddc5b150
954365,Total disclosed investment - Field: Privacy and security,constant 2021 US$,,2024-07-25 13:04:14,2024-07-30 06:18:26,,,2023-2023,6633,,$,"{""unit"": ""constant 2021 US$"", ""shortUnit"": ""$"", ""numDecimalPlaces"": 0}",0,,,disclosed_investment_summary__field_privacy_and_security,grapher/artificial_intelligence/2024-07-16/cset/cset#disclosed_investment_summary__field_privacy_and_security,"{""filters"": [{""name"": ""field"", ""value"": ""Privacy and security""}], ""originalName"": ""Total disclosed investment"", ""originalShortName"": ""disclosed_investment_summary""}",2,,,,,,,"Total disclosed investment between 2013-2023. Only includes private-market investment flows, such as venture capital; excludes all investment in publicly traded companies, such as the ""Big Tech"" firms. This data is expressed in US dollars, adjusted for inflation.",,"[""World aggregate does not include data for Micronesia, Tonga, Samoa, Kiribati, Fiji, Papua New Guinea, Palau, Tuvalu, Gibraltar, Jersey, Kosovo, Moldova, Isle of Man, Andorra, Montenegro, San Marino, Liechtenstein, Monaco, Vatican City, Afghanistan, Kyrgyzstan, Laos, Hong Kong, Bhutan, Brunei Darussalam, Maldives, Syria, North Korea, Myanmar, Timor-Leste, Nepal, Turkmenistan, Palestine, Yemen, Kuwait, Cape Verde, Equatorial Guinea, Swaziland, Namibia, Central African Republic (the), Angola, Ethiopia, Niger, Benin, Gabon, Gambia, Rwanda, Burkina Faso, São Tomé and Príncipe, Burundi, Guinea, Guinea-Bissau, Cameroon, Sierra Leone, Lesotho, Somalia, Chad, Liberia, Libya, South Sudan, Congo, Sudan, Malawi, Togo, Mali, Djibouti, Mauritania, Eritrea, Mozambique, Comoros, Antigua and Barbuda, Bolivia, Suriname, Nicaragua, Bahamas, Saint Vincent and the Grenadines, Grenada, Guyana, Haiti, Honduras, Cuba, Turks and Caicos Islands, Saint Lucia, and Dominica.""]","- Reporting a time series of AI investments in nominal prices (i.e., without adjusting for inflation) means it makes little sense to compare observations across time; it is therefore not very useful. To make comparisons across time possible, one has to take into account that prices change (e.g., there is inflation).
- It is not obvious how to adjust this time series for inflation, and we debated it at some length within our team.
- It would be straightforward to adjust the time series for price changes if we knew the prices of the specific goods and services that these investments purchased. This would make it possible to calculate a volume measure of AI investments, and it would tell us how much these investments bought. But such a metric is not available. While a comprehensive price index is not available, we know that the cost for some crucial AI technology has fallen rapidly in price.
- In the absence of a comprehensive price index that captures the price of AI-specific goods and services, one has to rely on one of the available metrics for the price of a bundle of goods and services. In the end we decided to use the US Consumer Price Index (CPI).
- The US CPI does not provide us with a volume measure of AI goods and services, but it does capture the opportunity costs of these investments. The inflation adjustment of this time series of AI investments therefore lets us understand the size of these investments relative to whatever else these sums of money could have purchased.",,,"{""note"": ""Data is expressed in constant 2021 US$. Inflation adjustment is based on the US Consumer Price Index (CPI). Data for 2022-2023 is incomplete.""}",int,[],8e869527235f14eb4250c3712c5b9525,973609a6e73ddf0f49758c6ce891af95