What new tasks will generative AI bring to humans?\u201d Acemoglu asks. \u201cI don\u2019t think we know that yet, and that\u2019s the question. What applications will really change the way we do things?\u201d<\/p>\n\n\n\n
What are the measurable effects of AI<\/strong>?<\/p>\n\n\n\n Since 1947, U.S. GDP has grown by an average of about 3% per year, and productivity has grown by about 2% per year. Some forecasts claim that AI will double that growth, or at least create a higher-than-usual growth trajectory. In contrast, in a paper published in the August issue of Economic Policy, \u201cThe Simple Macroeconomics of Artificial Intelligence,\u201d Acemoglu estimates that AI will increase GDP by \u201cmodestly\u201d between 1.1% and 1.6% over the next decade, and productivity by about 0.05% per year.<\/p>\n\n\n\n Acemoglu based his assessment on recent estimates of the number of jobs impacted by AI, including a 2023 study by researchers at OpenAI, OpenResearch, and the University of Pennsylvania, which found that about 20% of U.S. jobs could be affected byAI<\/strong> capabilities. A 2024 study by researchers at MIT\u2019s Center for the Future of Technology, the Productivity Institute, and IBM found that about 23% of computer vision tasks that could eventually be automated could be profitable over the next decade. Still more studies have put the average cost savings from AI at about 27%.<\/p>\n\n\n\n When it comes to productivity, \u201cI don\u2019t think we should underestimate a 0.5% increase over 10 years. It\u2019s better than zero,\u201d Acemoglu said. \u201cBut it\u2019s disappointing compared to the promises made by people in the industry and in the tech press.\u201d<\/p>\n\n\n\n To be sure, this is just an estimate, and many more AI<\/strong>applications are likely: As Acemoglu wrote in his paper, his calculations did not include using AI to predict the shapes of proteins, for which other academics subsequently won a Nobel Prize in October.<\/p>\n\n\n\n Other observers think that \u201creallocation\u201d of workers displaced by AI<\/strong> will generate additional growth and productivity beyond Acemoglu\u2019s estimates, though he thinks it\u2019s not significant. \u201cStarting from the actual distribution we have, reallocation generally yields only small benefits,\u201d Acemoglu says. \u201cThe immediate benefits are what matter.\u201d<\/p>\n\n\n\n \u201cI tried to write the paper in a very transparent way about what was included and what was not included,\u201d he adds. \u201cPeople can object and say that what I excluded was important or that the numbers for what I included were too low, and that\u2019s totally fine.\u201d<\/p>\n\n\n\n As Acemoglu and Johnson make clear, they favor technological innovations that increase worker productivity while keeping people employed, which should do a better job of sustaining economic growth.<\/p>\n\n\n\n But in Acemoglu\u2019s view, the point of generative AI<\/strong> is to mimic humans as a whole. This produces what he has for years called \u201cso-so technology,\u201d applications that perform at best only slightly better than humans but save companies money. Call center automation isn\u2019t always more efficient than humans; it just saves companies less money than workers do. AI applications that supplement workers seem generally to take a backseat to big tech companies.<\/p>\n\n\n\n \u201cI don\u2019t think complementary uses for AI<\/strong> will magically emerge unless industry invests a lot of effort and time,\u201d Acemoglu said.<\/p>\n\n\n\n What does history teach us about AI<\/strong>?<\/p>\n\n\n\n The fact that technology is often designed to replace workers is the focus of another recent paper by Acemoglu and Johnson, \u201cLearning from Ricardo and Thompson: Machinery and Labor in the Early Industrial Revolution\u2014and in the Age of AI,\u201d published in the August issue of the Annual Review of Economics.<\/p>\n\n\n\n The article discusses the current debate over AI, particularly the claim that even if technology replaces workers, the resulting growth will almost inevitably benefit society over time. Britain during the Industrial Revolution is sometimes cited as an example. But Acemoglu and Johnson argue that spreading the benefits of technology is not easy. In 19th-century Britain, they assert, it happened only after decades of social struggle and workers\u2019 action.<\/p>\n\n\n\n What is the optimal pace of innovation?<\/p>\n\n\n\n If technology helps promote economic growth, then rapid innovation would seem ideal because it would bring growth faster. But in another paper, \u201cRegulating Transformative Technologies,\u201d in the September issue of the American Economic Review: Insights, Acemoglu and MIT doctoral student Todd Lensman offer another view. If some technologies have both benefits and disadvantages, then it is better to adopt them at a more measured pace while mitigating those problems.<\/p>\n\n\n\n \u201cIf the social harms are large and proportional to the productivity of the new technology, then higher growth rates will lead to slower adoption,\u201d the authors write in the paper. Their model suggests that, ideally, adoption should start out slow and then gradually speed up over time.<\/p>\n\n\n\n \u201cMarket fundamentalism and technology fundamentalism might claim that you should always develop technology at the fastest pace,\u201d Acemoglu says. \u201cI don\u2019t think there is such a rule in economics. More thoughtfulness, especially about avoiding harms and pitfalls, is warranted.\u201d<\/p>\n\n\n\n The model is a response to trends over the past decade or so, in which many technologies were hyped as inevitable and welcomed for their disruptive nature. In contrast, Acemoglu and Lensman suggest that we can reasonably judge the trade-offs involved with a particular technology, and aim to stimulate more discussion about this.<\/p>\n\n\n\n How can we get to the right pace forAI<\/strong>adoption?<\/p>\n\n\n\n If the idea is to adopt technology more gradually, how should that be achieved?<\/p>\n\n\n\n First, Acemoglu said, \u201cgovernment regulation has a role to play.\u201d However, it\u2019s not clear what type of long-term guidelines for AI the U.S. or countries around the world might adopt.<\/p>\n\n\n\n Second, he added, if the \u201chype\u201d cycle around AI<\/strong> abates, then the rush to use AI \u201cwill naturally slow down.\u201d This scenario might be more likely than regulation if AI doesn\u2019t soon turn a profit for companies.<\/p>\n\n\n\n