“As artificial Intelligence (AI) reshapes culture, science, labor markets, and the aggregate economy, experts debate its value, risks, and how quickly it will integrate into everyday life,” noted “The Longitudinal Expert AI Panel,” a recently published report. “Leaders of AI companies forecast transformative AI systems that cure all diseases, replace whole classes of jobs, and supercharge GDP growth by the 2030s. Skeptics see small gains at best, with AI’s impact amounting to little more than a modest boost in productivity — if anything at all.”
“Despite these clashing narratives, there is little work systematically mapping the full spectrum of views among computer scientists, economists, technologists in the private sector, and the public,” the report added. “What do these groups believe about AI’s future capabilities, adoption, and effects? Why do they believe what they do, and what mechanisms support those beliefs? Prior surveys offer opinions, but rarely comprehensively quantify those opinions, hampering policy guidance and evaluation.”
The Longitudinal Expert AI Panel (LEAP) is a three-year project that aspires to fill this gap by analyzing the probabilistic forecasts of experts, historically accurate superforecasters, and the public. LEAP “aims to create the most reliable map of the future of AI and its impacts … by tracking the views of leading computer scientists, industry professionals, policy researchers, and economists on the trajectory of artificial intelligence.”
I first learned about LEAP in a recent article in The Economist, “A new project aims to predict how quickly AI will progress.” “Some of the grandest claims may be made with at least one eye on marketing,” wrote The Economist. “Still, getting a true sense of the probable speed of AI development is important,” said LEAP’s principal investigator Ezra Karger, — senior economist at the Federal Reserve Bank of Chicago and Director of Research at the Forecasting Research Institute. “When the spectrum of plausible outcomes includes an appreciable portion of white-collar tasks being automated, or a tenth of all electricity in America being used for AI training and deployment, good forecasts matter.”
Rather than assessing vague claims about concepts like AGI, LEAP aims to offer specific, testable hypotheses, such as: “When will self-driving cars account for 20% of American ride-hailing trips? What proportion of the country’s electricity will be used for AI by 2040? What will be the benchmark scores for open-source and proprietary AI models in 2025, 2027 and 2030?”
LEAP is guided by three key principles:
- Accountability: LEAP forecasts are detailed and verifiable, encouraging disciplined thinking and allowing us to track whose predictions prove most accurate.
- The wisdom of well-chosen crowds: LEAP emphasizes a diversity of perspectives from people at the top of their fields.
- Decision relevance: LEAP’s policy-relevant forecasts help decision-makers plan for likely futures.
Launched earlier this year, as of November of 2025 LEAP had completed three surveys on high-level predictions of AI progress. Almost 350 experts from many fields participated in these surveys. LEAP also included 60 highly accurate superforecasters — whose forecasts can be shown by statistical means and by their performance in forecasting tournaments to be consistently more accurate than those of other experts. It addition, it also included around 1,400 of especially engaged members of the general public.
Let me summarize a few of the key insights from these first LEAP surveys.
Experts expect sizable societal effects from AI by 2040.
- 18% of work hours will be assisted by generative AI in 2030, up from approximately 4% in November 2024.
- 7% of U.S. electricity consumption will be used for training and deploying AI systems in 2030, and 12%, close to double, in 2040.
- Global private AI investment will reach $260 billion by 2030, up from the $130 billion in 2024.
- 15% of adults will report using AI for companionship by 2030 up from 6% today. By 2040, the number doubles to 30% of adults.
Overall speed of AI progress. “By 2030, the average expert thinks that 23% of LEAP panelists will say the world most closely mirrors a rapid AI progress scenario, where AI writes Pulitzer Prize-worthy novels, collapses years-long research into days and weeks, outcompetes any human software engineer, and independently develops new cures for cancer. Conversely, the average expert believes that 28% of panelists will indicate that reality is closest to a slow-progress scenario, in which AI is a useful assisting technology but falls short of transformative impact.”
Comparison to previous historically significant technologies. “By 2040, the median expert predicts that the impact of AI will be comparable to a ‘technology of the century,’ akin to electricity or automobiles. Experts also give a 32% chance that AI will be at least as impactful as a ‘technology of the millennium,’ such as the printing press or the Industrial Revolution and just 15% predict that AI will be no more impactful than a ‘technology of the year’ like the VCR.”
The top and bottom quartile of experts express substantial uncertainty about the trajectory of AI
- While the median expert predicts substantial AI progress, and a sizable fraction of experts predict fast progress, other experts disagree.
- The top quartile of experts forecast that 50% of newly approved US drug sales in in 2040 will be from AI-discovered drugs, compared to a forecast of just 10% for the bottom quartile of experts.
- The top quartile of experts give a 81% forecast that AI will solve or substantially assist in solving a solution to a Millennium Prize Problem by 2040, compared to a 30% forecast from the bottom quartile of experts.
Median experts expect significantly less AI progress than the industry leaders of frontier AI companies like OpenAI, Anthropic, and xAI
- Industry leaders predict human-level AI by 2026-2029, while the expert panel indicates longer timelines for human and superhuman capabilities.
- By 2030, the average expert thinks that there is only a 23% chance that the world will closely mirror the (“rapid”) AI progress scenarios that matches some of these claims.
- Median experts give a 20% chance that AI will substantially assist in solving a highly complex mathematical problem by 2030.
- Median experts predict a 2% growth in white collar-employment by 2030, while some industry leaders predict at least a 10-20% overall unemployment by 2030.
Experts predict much faster AI progress than the general public
- Experts predict that by 2030, 18% of US work hours will be assisted by generative AI, while the general public predicts that only 10% will be assisted by generative AI.
- The median expert predicts that usage of autonomous vehicles will grow from a baseline of 0.27% of all U.S. rideshare trips in Q4 2024 to 20% by the end of 2030, compared to the 12% prediction from the general public.
- Experts predict that the number of AI-engaged research paper in physics, materials science and medicine will increase from 3% in 2022, to 30% in 2030, while the general public predicts that 20% of such research papers will be AI-engaged by 2030.
- Experts predict that by 2040, 25% of sales from newly approved US drugs will have been discovered by AI compared to 15% of the general public.
- On average, experts give a 63% chance that AI will turn out be at least as impactful as a ‘technology of the century’ — like electricity or automobiles — whereas the public gives this a 43% chance.
There are a few differences in prediction between superforrecasters and experts, — some quite large
- Superforecasters and expert groups generally predict similar futures, but where they disagree, superforecasters usually predict less progress than expert groups.
- For example, the median expert predicts that use of autonomous vehicles will grow from 0.27% of all U.S. rideshare trips in 2024 to 20% by the end of 2030, whereas the median superforecaster predicts a growth of only 8%.
- Superforecasters also predict less societal impact from AI and less AI-driven electricity use.
- Drug discovery is the only setting where superforecasters are more optimistic than experts: By 2040, experts, on average, predict that 25% of sales from recently approved U.S. drugs will be from AI-discovered drugs. Superforecasters predict 45%, almost double.
“Policymakers, nonprofit and business leaders, and other stakeholders routinely consult experts to base their decisions on the perspective of experts, especially when faced with new technologies and high levels of uncertainty,” said “The Longitudinal Expert AI Panel” (LEAP) report in conclusion. “With the launch of LEAP, we fill an important gap by both measuring the full range of expert opinions on AI capability developments and their impact, and by capturing the underlying reasoning and evidence that supports these beliefs.”
“LEAP will continue to explore important questions regarding the future of AI. Public, high-profile proclamations about the technology are not necessarily representative of expert opinion, and we will search for agreement and disagreement among experts, the general public, and professional forecasters.”