Inside the black box of AI: How AI models actually think | Emmanuel Ameisen (Anthropic)
Source
lennysnewsletter.com
Author
Lenny Rachitsky
Date
Why it matters
Explains interpretability findings that model internal planning and general concept representations are real, which challenges the 'next-tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → predictor' mental model engineers use when prompting and evaluating agents.
Terms in this piece · Glossary
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
alignment — The work of making AI systems actually pursue what their builders and users intend, rather than something subtly or dangerously different.
podcastOpenAI’s Head of ChatGPT: We’re entering a new era of AI (again) | Tibo Sottiaux
podcastHow we built Grok Bot in a month | Roman Ugarte (SpaceXAI)
podcastAI’s third era: the rise of persistent AI coworkers | Tara Seshan (OpenAI’s product lead)
podcastOpenAI’s Head of Design: This is the best time in history to be a designer | Ian Silber
