Researcher Nick Cammarata says AI models can be mapped
Nick Cammarata, a researcher, argues that AI systems are not opaque and that their reasoning can be charted. His essay has drawn wide attention online.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
A blog post of about 26,000 words by researcher Nick Cammarata has spread widely online. It argues that AI models are not indecipherable black boxes and that researchers can precisely map how they think. AI safety researchers have welcomed the argument enthusiastically.
Why it matters
The argument has been described as a major challenge to prevailing views in the AI field. It suggests that researchers may be able to understand AI systems far more precisely than previously assumed.
Coverage
- “We were looking at it wrong”: The explosive claim that’s upended the AI world (Will Douglas Heaven)
- Explosive Document Claims That AI Is Not What We Think (futurism.com)
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says