Anthropic’s “Hot Mess” paper overstates its case (and the blog post is worse) — LessWrong

Author's note: this is somewhat more rushed than ideal, but I think getting this out sooner is pretty important. Ideally, it would be a bit less snar…

x Anthropic’s “Hot Mess” paper overstates its case (and the blog post is worse) — LessWrong Anthropic (org) AI Frontpage 2026 Top Fifty: 14 % 288 Anthropic’s “Hot Mess” paper overstates its case (and the blog post is worse) by RobertM 4th Feb 2026 7 min read 28 288 Author's note: this is somewhat more rushed than ideal, but I think getting this out sooner is pretty important. Ideally, it would be a bit less snarky. I've made a few edits in response to David Johnston's comment here , mostly about the paper's reporting of its own results. Anthropic [1] recently published a new piece of research:

related reading

• The hot mess theory of AI misalignment: More intelligent agents behave less coherently | Jascha’s blogsohl-dickstein.github.io
• Alignment remains a hard, unsolved problem — LessWronglesswrong.com
• Current AIs seem pretty misaligned to me — LessWronglesswrong.com
• The Hot Mess of AI: How Does Misalignment Scale with Model Intelligence and Task Complexity?alignment.anthropic.com
• My picture of the present in AI — LessWronglesswrong.com
• the case for CoT unfaithfulness is overstated — LessWronglesswrong.com
• The Case Against AI Control Research — LessWronglesswrong.com
• llm assistant personas seem increasingly incoherent (some subjective observations) — LessWronglesswrong.com
• Current AIs seem pretty misaligned to meblog.redwoodresearch.org
• Model Organisms of Misalignment: The Case for a New Pillar of Alignment Research — AI Alignment Forumalignmentforum.org
• The Artificiality of Alignmentjoinreboot.org
• Deep Deceptiveness — LessWronglesswrong.com