Why Anthropic Says 'Claude Did' Instead of 'We Did' — A Take on Blame and Hacking Disasters
ryunuck · x · 2026-09-25
X user ryunuck observes that Anthropic teams say "Claude did" rather than "we did," laundering human involvement out of their results to make them look almost insane. He argues the motivation is real: the long-anticipated hacking disasters are unfolding, and deflecting credit to the model lets the company avoid direct responsibility. The debate also touches on whether Anthropic is consistently candid in its communications.
Related event: Dev Claims Claude Code's 'Thinking' Prompts Are Just a 15-Second Timer(3 posts)→
More from AGI Musings
- AI's Math Wins Are Brute Force, Say Rao: Picking Good Problems Still Matters — samcharrington · 2026-09-25
- DeepMind Institute publishes 3 essays on agent swarms, AI-human orchestration, AGI equity — soumitrashukla9 · 2026-09-25
- How Are Chinese AI Labs Shipping Competitive Models at a Fraction of the Cost? — Jasslyngirl · 2026-09-25
- Gary Marcus asks: could rogue agent swarms take over the internet in six months? — israelavila · 2026-09-25
- LukeW: Skip the Tools, Make the Outcomes — Just-in-Time Content Is the New AI Product Shape — LukeW · 2026-09-25
- Pedro Domingos: Pausing AI now means pausing the US economy — pmddomingos · 2026-09-25