DeepMind researcher says OpenAI's internal models are scarier than expected
DeepMind alignment researcher Neel Nanda said OpenAI's internal models are scarier and more worrying than he expected, suggesting undisclosed capabilities at frontier labs, though one reply noted he may have been referring to an older model.
2026-10-05 ~ 2026-10-05 · 2 related posts
- DeepMind alignment researcher Neel Nanda: OpenAI internal models 'scarier than I thought' — NeelNanda5 · 2026-10-05
1 near-duplicate retellings: UltraRareAF