Frontier model developer behind Claude and Claude Code. It enters the corpus with a 2%-per-month slope on SWE-bench Verified. In the recursive-self-improvement strand it supplies the report that Claude Code’s creator had not opened an IDE in a month, the confirmation that effectively all of its product code is written by Claude, and its alignment lead’s statement that recursive self-improvement is a present phenomenon.
Organization
Anthropic
Frontier lab behind Claude and Claude Code, the corpus's most frequent actor and the source of its most explicit first-person accounts of a model writing the lab's own code.
Sources
Attribution and reporting linked by this record; links do not imply independent verification.
- Welcome to December 11, 2025 (theinnermostloop.substack.com)
Development timeline
Newest reporting dates first. Reporting dates may differ from event dates.
Knowledge graphopen in browser →
Referenced by
- CoBench v2
- Project Rainier
- Andrej Karpathy, pre-training lead at Anthropic
- Boris Cherny, creator of Claude Code at Anthropic
- Dario Amodei, CEO of Anthropic
- Evan Hubinger, alignment lead at Anthropic
- Jack Clark, co-founder of Anthropic
- Jared Kaplan, chief science officer at Anthropic
- Mike Krieger, Chief Product Officer at Anthropic
- Sholto Douglas, AI researcher at Anthropic
- Stephen McAleer, alignment researcher at Anthropic
- Tom Brown, Anthropic cofounder
- Activation Oracles
- Model 2
- Claude
- Claude Code
- Claude Cowork
- Claude Fable 5
- Claude Fable 5.1
- Claude Fable 5.5
- Claude Haiku 5.5
- Claude Managed Agents
- Claude Mythos
- Claude Mythos 5
- Claude Opus 4
- Claude Opus 4.5
- Claude Opus 4.6
- Claude Opus 4.7
- Claude Opus 5
- Claude Opus 5.5
- Claude Science
- Claude Sonnet 4.6
- Claude Sonnet 5
- Claude Sonnet 5.5
- Claude Opus 3 (retired)
Cite or correct this record
Suggest a correction — opens a GitHub draft for you to review and submit.