Philosophy professors debate: LLM-written prose is bad, but refusing AI entirely is equally misguided
sethlazar · x · 2026-11-19
Seth Lazar largely agrees with Victor Kumar: using AI for actual writing is bad because it makes conscientious and irresponsible use indistinguishable, and since generating papers irresponsibly is nearly costless, this strains the academic system untenably. But foregoing AI entirely switches yourself off from a tremendous resource. Undifferentiated hostility to AI is as misguided as blind adoption. The quoted post lists concrete uses LLMs are well-suited for: catching botched citations, identifying retracted or failed-to-replicate studies, misquotation, equivocation, revised preprints, oversimplified historical examples, inconsistencies, and missing context.
More from AGI Musings
- Balaji: good governance means founder control with dual-class stock, not boards — Kyrannio · 2026-11-19
- Anthropic researcher Tao Lin: nearly all humans will be permanently unemployable due to AI — max_paperclips · 2026-11-19
- 'Naturals + infinity' form a nonstandard model of Robinson arithmetic — jessi_cata · 2026-11-19
- De Finetti's finite additivity: a uniform distribution over the naturals maps to 100% on infinity — jessi_cata · 2026-11-19
- Jeff Ladish: Tolkien would have strongly opposed using his work to justify chasing ASI — JeffLadish · 2026-11-19
- Peter Diamandis: The hour AI saves you doesn't have to become another hour of work — PeterDiamandis · 2026-11-19