Suivre
- X@ch402
- Site webcolah.github.io
- Wikipédiaen.wikipedia.org
Chris Olah
Cofondateur d’Anthropic et responsable de la recherche en interprétabilité, il avertit que personne ne comprend comment fonctionnent les réseaux de neurones de pointe, que les incitations des laboratoires entrent en conflit avec le fait de bien agir, et que l’IA pourrait remplacer le travail à très grande échelle.
Fonctions
- 2015 – 2018Research Scientist (from intern), Google BrainCo-founded the journal Distill in 2017.Source →
- 2018 – 2020Lead, interpretability team, OpenAISource →
- 2021 – aujourd’huiCo-founder and Interpretability Research Lead, AnthropicLeft OpenAI with the group that founded Anthropic; leads the mechanistic interpretability team.Source →
Au registre
juil. 2026 · Lettre ouverte · Pacing the Frontier
There is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems.
Co-founder and Interpretability Research Lead, Anthropic à l’époque
Signed as "Co-Founder & Interpretability Research Lead, Anthropic".
25 mai 2026 · Discours · Vatican, presentation of the encyclical Magnifica humanitas
Every frontier AI lab—including Anthropic—operates inside a set of incentives and constraints that can sometimes conflict with doing the right thing. … We need informed critics who will tell the labs when we are failing. We need moral voices that the incentives cannot bend.
Co-founder and Interpretability Research Lead, Anthropic à l’époque
Spoke beside Pope Leo XIV at the encyclical's release; also warned that large-scale labour displacement would make support for the displaced "a moral imperative of historic proportions".
Anthropic co-founder Chris Olah's remarks on Pope Leo XIV's encyclical "Magnifica humanitas" →
5 sept. 2024 · Entretien · TIME
If we could really understand these systems, and this would require a lot of progress, we might be able to go and say when these models are actually safe. Or whether they just appear safe.
Co-founder and Interpretability Research Lead, Anthropic à l’époque
30 mai 2023 · Déclaration · Center for AI Safety
Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks such as pandemics and nuclear war.
Co-founder and Interpretability Research Lead, Anthropic à l’époque
Listed as "Chris Olah, Co-Founder, Anthropic" (confirmed in the full signatory list).
août 2021 · Podcast · 80,000 Hours Podcast
We don't know how they're doing the things they do, or why they're doing them, and really don't know how to reason about how they might behave in other unanticipated situations.
Co-founder and Interpretability Research Lead, Anthropic à l’époque
On the risk of deploying systems whose internals are not understood.
Chris Olah on what the hell is going on inside neural networks →