AI Research Leaders Urge Policymakers to Examine Automated AI Research
The researchers warned that automating AI development could compress years of progress into months or less and accelerate technological change beyond human understanding.
Research leaders at OpenAI, Anthropic, Meta Platforms and Microsoft urged policymakers Monday to examine how extensively their companies have automated artificial intelligence research, warning that faster development could outpace human understanding.
The researchers said automated AI development could trigger an “intelligence explosion,” compressing advances that might otherwise take years into months or less. That acceleration could increase the pace of technological change beyond people’s ability to understand it.
Writing personally, the authors called for policymakers to require greater transparency from companies developing AI models. The proposed transparency would help regulators assess how much of the research and development process has been automated.
The group includes OpenAI Chief Scientist Jakub Pachocki, Anthropic co-founder Jack Clark, Microsoft Chief Scientific Officer Eric Horvitz and Meta Vice President of Artificial Intelligence Research Dawn Song.
Their call comes amid broader debate over regulating increasingly advanced AI systems and automated AI research.
