AI may unleash an unprecedented acceleration in human progress, potentially packing the equivalent of 10,000 years of technological advancement into a mere 25 years, according to AI risk assessment expert Ajeya Cotra.
Cotra, who works with METR, a non-profit focused on evaluating when AI systems might present catastrophic risks to society, shared her striking forecast in a recent episode of the 80,000 Hours podcast hosted by Rob Wiblin. She warned that many observers, including economists, may be significantly underestimating AI’s transformative potential.
“I think that there’s a pretty good chance that by 2050 the world will look as different from today as today does from the hunter-gatherer era,” Cotra stated. “It’s like 10,000 years of progress rather than 25 years of progress driven by AI automating all intellectual activity,” she added.
AI expert expects 10,000 years of tech progress
Cotra’s argument centres on the possibility of AI systems automating not only routine tasks but core drivers of innovation – software development, scientific research, and even aspects of physical production. Once AI reaches the point of outperforming top human experts in computer-based intellectual work, something she anticipates could emerge in the early 2030s, progress could enter explosive feedback loops.
At that stage, AI would accelerate its own improvement, designing superior models far faster than human researchers could. Combined with advances in robotics and manufacturing, this could enable AI to construct the physical infrastructure, such as advanced chips and massive data centres, which are needed to support even more powerful systems.
Bigger changes to society expected
Cotra contrasts this scenario with more conventional economic forecasts, which often project AI-driven growth resembling the gradual acceleration seen during the Industrial Revolution over the past century. Instead, she envisions a compounding effect where automation of intellectual labour leads to radically faster innovation across domains, from software engineering to fields like virology.
The implications are greater too – by mid-century, society could undergo changes on a scale comparable to the entire span of human civilisation since the hunter-gatherer era.
Cotra isn’t the only one who predicts such drastic changes in the coming years. Elon Musk, head of xAI, has predicted that AI could surpass individual human intelligence by 2026 or 2027 and exceed all of humanity collectively by 2030–2031, with robotics amplifying the economic explosion. Demis Hassabis, Nobel laureate and CEO of Google DeepMind, has described AI’s potential impact as “10 times bigger than the Industrial Revolution—and maybe 10 times faster.”