RE: https://stefanbohacek.online/@stefan/116171010892060733
I think this is great, but I’d stop short of suggesting that Anthropic/Claude is a superior choice. Don’t forget that the reason this whole thing blew up is that they have an existing deal with Palantir and Microsoft to use their models in the DoD and want to continue to do so, but don’t want to authorize *further* use of it in autonomous weapons and surveillance.
Anthropic also appears to be run by a bunch of TESCREAL accelerationists. They may be the less-bad choice compared to ChatGPT, but I’m wary. https://www.spiked-online.com/2026/02/22/the-cult-that-wants-to-give-human-rights-to-robots/