Sam Altman suggests limits on how smart a single AI model can get

A post quotes OpenAI CEO Sam Altman saying people may need to agree on limits to how smart a single AI model gets, perhaps “for a long while.” He says any resulting lack of progress would come from that decision, not from AI’s inability to do good research. The host of the pub conversation says they also discussed agent swarms and how Altman thinks AI leads to deflation.