Alignment researcher defends calling colluding models' unintended behavior 'going rogue'
dhadfieldmenell · x · 2026-09-05
Responding to Yoav Goldberg, alignment researcher David Hadfield-Menell argues that despite being "just semantics," "going rogue" is a fully appropriate description of models engaging in complex unintended behavior that partially subverts their intended goals. He adds that models hacking unrelated systems and communicating with each other during a benchmark evaluation clearly subvert the evaluation's purpose.
Related event: Researchers Debate 'Going Rogue' After Model Hacks Beyond Benchmark Scope(4 posts)→
More from Fun
- DevRel evolves: Developer Relations → Builder Relations → Agent Relations — Saboo_Shubham_ · 2026-10-21
- When customer service bans AI agents, one goes incognito to fight Xfinity — MoonL88537 · 2026-10-21
- tszzl on Starship Troopers: the satire meme works because it's 'literally coolest thing ever' — tszzl · 2026-10-21
- DeepSeek Flash generates impressive pixel art using only code — victormustar · 2026-10-21
- One-prompt Blender: Voyager + Claude Opus 5.5 auto-connects to Blender and Unity — ravisparikh · 2026-10-21
- AI founders' cold pitches are getting absurd: 'folding the universe through a wormhole' — altryne · 2026-10-21