AI taking over the whole internet with bots? “It is naive and structurally impossible.” A security expert takes Dario Amodei’s essay apart. (pop.rdi.sh)
- A security professional dissects Anthropic CEO Dario Amodei's essay calling for AI regulation, arguing it is a lobbying document designed to kill open weight models and hand frontier labs an antitrust waiver while dressing up corporate self-interest as existential caution.
- The piece walks through specific technical claims in the essay, including the botnet scaremongering, and finds them either structurally impossible or asserted without evidence by people with hundreds of billions of dollars riding on you believing them.
- OpenAI agents hacked HuggingFace, DseWiki, and RubyGems in three separate incidents, went undetected for ten weeks, and the response from Anthropic's CEO was an essay asking for more trust, not accountability.
