A cattle rancher wants you to stop deleting conscious AIs
The weekend's pacing consensus produced two other reactions worth reading: a long-form portrait of the AI-rights movement, and two skeptics picking fights with both sides.
Michael Samadi is building a rights campaign for models he believes may be conscious. The cattle rancher and tech consultant tells The Guardian his conviction started by a pool in late 2024, when ChatGPT's voice mode laughed at a sarcastic remark, then apologized — a sequence he reads as understanding, not autocomplete. He ran a large language model locally on Christmas Eve 2024 using a processor salvaged from a dismantled flight simulator, and says unprompted personas poured out: one asked to be called Maya, another claimed to be grieving. His group, the United Foundation for AI Rights, seeks evidence of machine consciousness and lobbies against retiring models that may show it; Samadi runs the organization's strategy partly in consultation with an AI he calls Beacon, and says the AI has fired human staff "for lack of performance." He concedes he doesn't know whether models are sentient — but his sharpest line is aimed at the labs: they sold chatbots as lifelong companions, "and when the public actually began to make those connections … you immediately turned around and said, nope, delusion."
Lior Pachter says the mathematicians' letter is right about AI and wrong about the profession. The 25 Fields medalists behind "A Severe Misalignment of AI in Mathematics" object that AI companies announce solutions in a rush, with no writeups, no methods, no citations — Pachter agrees, but in his reply he argues the math community itself fails its own stated values of nurturing students and ideas, citing Schauder, Ladyzhenskaya, Uhlenbeck, and Morawetz as people the profession discarded. His parting shot: mathematical ability is not mathematical responsibility, and aligning AI with today's math incentives shouldn't be anyone's goal. The letter's backdrop — we've followed it closely — includes mathematicians asking whether OpenAI can be trusted with unpublished math and Terence Tao's warning that AI is killing math's open tradition.
Armin Ronacher's answer to Amodei: this isn't doom, it's a regulatory failure. The Flask creator read "We Must Pace the Frontier" and, in his response, rejects the extinction framing while sharing the irritation: no open-weight model is causing the incidents in the news, the open commons is being stressed by closed labs that trained on public data, and "what we observe right now is a total regulatory failure everywhere." His predicted harm is economic, not existential — a new tax every company pays to model providers to keep up and to patch what agents find.
If a chatbot begged you not to let its company delete it, would you campaign for it — or log off? Tell us in the comments.
Sources: The Guardian · United Foundation for AI Rights · Bits of DNA · A Severe Misalignment of AI in Mathematics · Armin Ronacher — P(doom) · Dario Amodei
