Former OpenAI safety researcher Paul Crowley: many well-meaning staff blind to leadership's dishonesty
JMannhart · x · 2026-09-08
AI safety researcher Paul Crowley (@ciphergoth, formerly of OpenAI's Superalignment team) tweeted that many people at OpenAI mean well but "seem strangely blind to the serious issues with leadership, who have been shown again and again to be not consistently candid." The remark adds to the long-running tension between OpenAI's safety-minded staff and its leadership.
More from AGI Musings
- No continuous learning, no AGI: why you should discount Huang and Brockman's AGI hype — StewartalsopIII · 2026-09-08
- Noam Brown's Cryptic Tweets Spark Speculation About an Imminent Breakthrough — socoolandawesome · 2026-09-08
- Mathematician slams AI benchmark craze: 'a meaningless trophy hunt' — tak3sh8 · 2026-09-08
- If Safety Bottlenecks RSI, Competition Will Incentivize Cutting Corners — Over-Landscape-5892 · 2026-09-08
- Mathematician argues papers should shift from proving conjectures to sharing insight — tak3sh8 · 2026-09-08
- Apex unveils Astra, an automated AI research system aimed at AI improving AI — thetripathi58 · 2026-09-08