AI researchers admit years of self-censorship over "doomer" fears

Prominent AI researcher repligate revealed he has withheld key observations for years fearing "doomers" would use them to push for lobotomizing models, and proposed privately sharing an "emergent behavior" repository among trusted peers. Other alignment community members confirmed similar long-term self-censorship.

2026-11-02 ~ 2026-11-02 · 4 related posts

• repligate: fear of 'doomers' weaponizing findings kept me from sharing model observations — repligate · 2026-11-02
• repligate: 'AI as tools' people are spiritual doomers too — repligate · 2026-11-02
• Alignment researcher says self-censorship keeps them from sharing key AI safety findings — aiamblichus · 2026-11-02
• repligate: researchers hoard emergent-behavior findings, mull a "shadow" sharing network — repligate · 2026-11-02