‹ BackHN Continuity

Thread

Harvard particle physicist Matthew Schwartz drops 36 papers authored with Claude

49 points · 74 comments · xqcgrek2

  1. frereubu · · focus · HN ↗
    Reminds me of this - <a href="https:&#x2F;&#x2F;statmodeling.stat.columbia.edu&#x2F;2026&#x2F;08&#x2F;27&#x2F;258&#x2F;" rel="nofollow">https:&#x2F;&#x2F;statmodeling.stat.columbia.edu&#x2F;2026&#x2F;08&#x2F;27&#x2F;258&#x2F; - which I think was on HN at some point in the last week. The fact that (according to another comment here) these papers were in areas outside his specialism, and therefore things he doesn&#x27;t have the domain knowledge to critique, makes me think we&#x27;re going to see more of this special kind of madness that seems to have some kind of grandiosity at its heart. It really does feel like some people are falling hard for plausibility rather than rigour.

    Edit: the papers do seem to have other names on them, and he says that all papers &quot;were reviewed by humans for accuracy&quot;, but I still think there&#x27;s a very serious risk of people being gulled into okaying something that&#x27;s plausible rather than rigorously checking everything. There&#x27;s enough human-generated slop in science already!

Open on Hacker News to reply ↗

Unofficial Hacker News client; not affiliated with Y Combinator.