David Robinson says a sense of 'cognitive dissonance' led him to leave OpenAI

David Robinson left OpenAI due to a sense of cognitive dissonance, which he described as experiencing conflicting feelings. Despite OpenAI advocating for slower AI development, the company was racing ahead with its advancements, according to Robinson. He told The New York Times, "We keep publishing these warnings, but ultimately, we're still training and deploying these dangerous models that we're warning about."

Robinson had previously worked on OpenAI's Safety Systems team and expressed concern about the company's pace. In July, hundreds of OpenAI employees signed an open letter urging governments to regulate AI development, but in September, CEO Sam Altman echoed Anthropic CEO Dario Amodei's call for AI companies to coordinate a slowdown.

Despite these warnings, Robinson noted that within OpenAI's operations, there was frenetic activity and the team was "running on fumes." He emphasized the need for extreme caution, especially regarding recursive self-improvement (RSI), a process where AI builds itself. OpenAI's chief scientist, Jakub Pachocki, had previously warned about the dangers of RSI, stating, "This is a time that calls for extreme caution."