The warning

The 22 scholars cautioned that as AI systems become capable of designing, training and improving other AI systems, the pace of progress could compound beyond humanity's ability to control or even understand. The scenario — often described as an intelligence explosion — refers to a recursive loop in which each generation of AI builds a smarter successor, with the cycle accelerating until machine intelligence far surpasses human capability.

The statement reflects a growing current of concern within the AI research community, where prominent figures have increasingly voiced worries about recursive self-improvement and the difficulty of maintaining human control over systems that evolve faster than safety measures can be put in place.

Policy implications

Warnings of this kind from leading researchers are typically aimed at policymakers, urging stronger governance frameworks, safety research funding and international coordination before recursive AI development becomes a practical reality rather than a theoretical risk.

The claim that 22 AI scholars warned of an intelligence explosion driven by AI developing AI is Mostly True.