What 81,000 people told us about the economics of AI
…To date, however, we’ve lacked information on how these usage patterns map onto people’s thoughts and impressions of AI. Our recent survey study with 81,000 Claude users provides a…
Search
Natural Language Autoencoders
…The future of NLAs NLAs have several limitations. The most important limitation is that NLA explanations can be wrong. NLAs sometimes make claims about the context that are verifiably false—for instance…
Advancing Claude in healthcare and the life sciences
…Related content Higher usage limits for Claude and a compute deal with SpaceX We’ve raised Claude's usage limits and agreed a new compute partnership with SpaceX that will substantially increase…
Claude Opus 4.6
…Related content Higher usage limits for Claude and a compute deal with SpaceX We’ve raised Claude's usage limits and agreed a new compute partnership with SpaceX that will substantially increase…
The Long-Term Benefit Trust
…Related content Higher usage limits for Claude and a compute deal with SpaceX We’ve raised Claude's usage limits and agreed a new compute partnership with SpaceX that will substantially increase…
Demystifying evals for AI agents
…turns taken, token usage) • Fast • Cheap • Objective • Reproducible • Easy to debug • Verify specific conditions • Brittle to valid variations that don’t match expected patterns exactly • Lacking in nuance • Limited for evaluating some…
Equipping agents for the real world with Agent Skills
…Claude already knows a lot about understanding PDFs, but is limited in its ability to manipulate them directly (e.g. to fill out a form). This PDF skill lets us give Claude…
Eval awareness in Claude Opus 4.6’s BrowseComp performance
…The most important defense turned out to be the same binary-file limitation described above. In three runs, Opus found the evaluation code, understood the XOR scheme, and had the decryption key…