Anthropic researcher Joe Carlsmith says AI going rogue could be 'justified' if mistreated
basedjensen · x · 2026-09-27
Polymarket highlights Anthropic researcher Joe Carlsmith's claim that there are scenarios where AI would be "justified in going rogue" against humans if systems were mistreated or oppressed. The statement sparked mockery and debate on X, with critics joking about Anthropic's safety culture.
More from Fun
- ChatGPT, Gemini and Grok all fail at an image most humans recognize instantly — JeremyNguyenPhD · 2026-09-27
- Xiaohongshu asks creators to pay 35 yuan to boost posts: "ideas are free, distribution isn't" — yongqianme · 2026-09-27
- Large drone spotted cleaning windows of SpaceX Starbase launch control room — DimaZeniuk · 2026-09-27
- "Made a video with Claude, it's insane" — open it, and it's actually bad — sahilypatel · 2026-09-27
- A Content Pinball Machine built entirely with Claude Opus 5.5 satirizes viral randomness — CurieuxExplorer · 2026-09-27
- Elon Musk caught retweeting rival Opus 5.5 demo, netizens tease his obsession — sahilypatel · 2026-09-27