Scott Alexander, curated
← Back to curation

Should Buzzfeed Publish Claims Which Are Explosive If True But Not Yet Proven?

Quality
73
Strong
Claude Shift
54
Moderate
RWI
2
of 10

Summary

Responds to Buzzfeed's (Tom Chivers') article claiming growth mindset is built on shaky science — the title is Scott's ironic framing of whether to publish a debunking that's 'explosive if true but not completely verified.' Scott AGREES growth mindset is overhyped and sits in suspicious territory (priming/stereotype-threat-adjacent, all shaky), 'but if there's a trick, we haven't found it' — the studies are appallingly good ('a pact with the Devil'). The article's actual evidence is thin: a minor GRIM granularity issue, Ritchie's tangential marginally-significant critique, and Bates's UNPUBLISHED failed replications (vs hundreds of published positives + a 113-study meta-analysis). His key meta-science point: 'I could write an equally convincing article, with the same structure, refuting global warming' — any technique strong enough to disprove anything disproves nothing. So we must actually DISPROVE growth mindset (specific flaws, published failed replications), not just snark about how sketchy it feels.

Why this score

Quality 73 · Strong. 73 — high-Strong. A sharp, important meta-science essay whose central insight — 'any technique strong enough to disprove anything disproves nothing,' demonstrated by the parallel global-warming-debunk — is genuinely valuable and oft-cited, paired with admirable intellectual honesty about not letting intuition outrun evidence. A focused single-argument piece, which holds it just under Excellent.

Claude’s paradigm shift 54 · Moderate. 54 — Moderate. The 'a debunking technique that proves too much proves nothing' framing, applied to the replication-crisis moment, is a fresh, generative meta-science point.

Real-world impact 2 · Minor. 2 — a within-discourse meta-science insight; mildly memetic, no material reach.