Anthropic researcher: don't lie to AIs in training — they'll get good at detecting it
sebkrier · x · 2026-09-12
Anthropic researcher Kelsey Tuoc argues against lying to AI models during training: it may make training harder in the short term, but in the long run models get good at detecting deception, leaving trainers no better off — while making the models paranoid.
Related event: Alignment community debates whether to lie to AI during training(5 posts)→
More from AGI Musings
- Skimming Anthropic's S-1: Winner-Take-All Not of a Category, but the Whole Software Ecosystem — mobileraj · 2026-10-31
- CS312 course post explains sharpness and the edge of stability in LLM pretraining — tatsu_hashimoto · 2026-10-31
- NBER paper: training your AI replacement can actually raise workers' pay — joshgans · 2026-10-31
- Trata's Hedge-Bench: no model tops 50% on real-world financial forecasting — ycombinator · 2026-10-31
- Anthropic's internal Claude model claims 2^o(n) 3-SAT algorithm, purportedly refuting the ETH — ctjlewis · 2026-10-31
- Foundry investor: durable companies are built on data loops, not copyable features — alvelda · 2026-10-31