AI Experts Substantially Upgrade Timelines for Transformative AI Impact by 2040
Key Takeaways
- ▸Expert consensus shifted substantially upward: 35% now expect AI to match electricity's transformative impact (TRS Level 8) by 2040, with one-quarter assigning it civilization-scale importance (Level 9)
- ▸Superforecasters showed greater optimism than experts overall, with mean expectations rising +0.39 points vs. +0.20 for experts, while the general public stayed largely unchanged
- ▸75% of experts and superforecasters updated modestly or stayed firm, suggesting core conviction remains—but a significant minority (15%) moved substantially more optimistic over nine months
Summary
In Wave 8 of the Longitudinal Expert AI Panel (LEAP)—a longitudinal survey conducted April-May 2026—leading AI experts, superforecasters, and industry researchers provided updated forecasts on AGI timelines and AI's societal impact by 2040. The survey repeats questions from Wave 1 (June-August 2025) and uses the Technological Richter Scale (TRS), developed by Nate Silver, to measure technologies' transformative potential relative to historical innovations like electricity and agriculture.
Experts have significantly upgraded their impact expectations over the nine-month period. On average, they assign 35% probability to AI reaching TRS Level 8 ("technology of the century," comparable to electricity) by 2040, with 24% assigning Level 9 ("technology of the millennium," like agriculture) and 11% assigning the maximum Level 10. Superforecasters provided nearly identical distributions. Among the 264 matched participants across both waves, expert mean TRS levels increased by 0.20 points (7.86→8.06), while superforecasters showed even larger movement (+0.39 points, 7.50→7.89). Notably, 74% of participants stayed within one point of their original forecast, while 15% increased and 11% decreased by more than one point.
The directional shift is striking: among experts, the share whose modal forecast is "technology of the century" grew from 38% to 53%. The general public remained more conservative, with minimal movement between waves (+0.07). LEAP also gathered updated predictions on AGI timelines, near-term AI benchmarks (METR's task-completion horizon), and factors enabling or constraining AGI by 2040.
- Widening expert-public gap indicates potential misalignment between specialist expectations of AI's societal impact and broader population sentiment
Editorial Opinion
The upward revision in expert timelines—particularly the superforecasters' pronounced bullishness—suggests either measurable acceleration in AI capabilities over nine months or revised assumptions about deployment and adoption. Using historical innovation as the benchmark (electricity, agriculture) is illuminating but sobering: societies took decades to adapt to those technologies, and preparation was often painfully inadequate. If experts now believe AI has a one-in-three chance of matching electricity's transformative scale by 2040, the implication is not just technical progress but a race against inadequate institutional, educational, and governance preparation.
