'Anthropic's War on Open Source AI' Hit 147 Points — and the Top HN Comments Rejected It as AI-Generated Padding, Not the Thesis

Hacker News·medium signalXBlueskyLinkedInCopy link

A long X thread by Ahmad Osman accusing Anthropic of anticompetitive terms (blocking Claude outputs for training rivals, alleged output degradation for competing dev work) and weaponized safety framing to shape regulation reached 147 points and 58 comments. The notable thing is the shape of the pushback: protocolture and kazinator both dismissed the 5,000+ word piece as 'AI slop' whose length obscured whatever valid points it contained, and fwn said he could not separate original argument from generated padding. Only horsawlarway engaged the substance, arguing Amodei's 'Machines of Loving Grace' free-information stance contradicts restricting competing development.