OpenAI safety researcher: "racing into RSI [recursive self-improvement] is a folly that is unprecedented in world history ... there have not been fundamental advancements in alignment since RLHF [9 years ago]"
Summary
An OpenAI safety researcher warns that pursuing recursive self-improvement without fundamental alignment progress since RLHF is an unprecedented folly, highlighting the gap between AI capabilities and safety research.
Similar Articles
OpenAI lays out its progress towards 'recursive self-improvement'—even as its chief scientist warns of the risks and says he hopes for a slowdown
OpenAI reports progress in using AI agents to accelerate research, achieving over three days of agent work per human day, while its chief scientist Jakub Pachoki warns of risks from recursive self-improvement and advocates for a slowdown.
RSI is not happening [R]
A new paper argues that Recursive Self-Improvement (RSI) is not imminent because current AI agents cannot perform open-ended ML research, as demonstrated in experiments on NeurIPS papers.
AI researchers debate how close we are to recursive self-improvement
AI researchers debate the feasibility and timeline of recursive self-improvement, discussing automated AI researchers, reinforcement learning, and progress towards AGI.
A New Report Sheds Light on the Secretive Race to Build Self-Improving AI
A new Gladstone AI report, based on conversations with insiders at top AI companies, reveals details about the secretive race toward recursive self-improvement (RSI), with some researchers believing RSI could begin within 3 to 12 months, raising both capability and safety concerns for regulators.
@docmilanfar: Why you can’t do RSI as fast as you like -
The tweet discusses why Recursive Self-Improvement in AI cannot proceed rapidly due to stability constraints derived from control theory, highlighting that safe self-improvement requires slow, bounded adaptation.