williawa comments on Anthropic’s “Hot Mess” paper overstates its case (and the blog post is worse)

[williawa]

But you agree the Anthropic post does not demonstrate, or even really provide meaningful evidence for that, right?