在美国经济中衡量人工智能产生的概念和挑战.pdf

1、IMF IMF|Statistics1Concepts and Challenges of Measuring Production of Artificial Intelligence in the US EconomyNOVEMBER 21,2024Tina Highfill(Dave Wasshausen,Greg Prunchak)National Economic Accounts/Senior Research Economist/US Bureau of Economic Analysis IMF IMF|Statistics2OutlineOutlineConcepts and

2、 challenges of measuring AI production How production of AI software is currently reflected in the Supply and Use Tables(SUTs)Possibilities for measuring AI production using a thematic satellite account frameworkIMF IMF|Statistics3Why Measure Production of AI?Why Measure Production of AI?Current AI

3、research mostly focuses on uses of AI“Augmented Intelligence:The Effects of AI on Productivity and Work Practices”(Raymond,Brynjolfsson,and Li 2024)IMF IMF|Statistics4Why Measure Production of AI?Why Measure Production of AI?Current AI research mostly focuses on uses of AIUS federal government actio

4、ns to encourage domestic AI manufacturing and research 2022 CHIPS and Science Act(CHIPS=“Creating Helpful Incentives to Produce Semiconductors”)2023 Executive Order on the Safe,Secure,and Trustworthy Development and Use of Artificial Intelligence“U.S.companies lead the world in driving AI forward,do

5、minating the design of AI chips and the development of Large Language Models.But we dont manufacture or package any of the leading-edge AI chips needed to fuel the innovation ecosystem and power our most critical defense systems.We need more talent development,R&D,and manufacturing to take place in

6、America.”Gina Raimondo,US Secretary of Commerce(February 2024)IMF IMF|Statistics5Production of AI Software in Production of AI Software in the Supply and Use Tables the Supply and Use Tables IMF IMF|Statistics6Production of ForProduction of For-sale AI Software in the SUTs sale AI Software in the SU