AI researcher torches hit piece: all major chatbots give near-identical answers to the same prompt
KelseyTuoc · x · 2026-10-20
AI researcher Kelsey Tuoc published a multi-post rebuttal of journalist @aaronsibarium's article on Claude's handling of sensitive questions (e.g., returning search results to adults posing as 17-year-olds), calling it malpractice.
Her key points:
- Every major model (Grok, ChatGPT, Gemini, etc.) gives basically identical answers in the scenario the piece advertises, so singling out Claude is misleading reporting.
- Real divergences do exist — Claude, for instance, cares more about animal cruelty — or one could write about how all AIs are generic 2010s liberals; the author did neither.
- Anthropic's under-18 ban is enforced via account-behavior surveillance, which wouldn't catch adults lying about being teenagers.
- She says the piece's potentially interesting parts about Claude's Constitution are sabotaged by clickbait framing and argues it should be taken down.
More from Fun
- Domingos jokes: I foresaw Nvidia's AI boom in my 2015 book — but never bought the stock — pmddomingos · 2026-10-20
- Homelab fans spend $6,000 and 14 weekends on ZFS just to dodge Google's $1.99 storage fee — rachittshah · 2026-10-20
- AI works for 40 seconds then replies "This response couldn't load" — Tasty_Chapter_6742 · 2026-10-20
- Founder runs 12 parallel Claude agents for 6 hours to reverse-engineer Supabase, open-sources it — dshukertjr · 2026-10-20
- "Alea Iacta Est," 2026 Edition: jax.random.key(42) — fhuszar · 2026-10-20
- Agent hits level 49 in World of Warcraft with a live map dashboard — max_paperclips · 2026-10-20