historical marker: in October 2026, when two critical security vulnerabilities are found, everyone simply assumes that they were found with AI, and (at least in the PHP case, no confirmation about the other one) they were correct

Show Post

My condolences. ... I've been waiting to hear your reaction ever since hearing the news.

Show Post

Show Post

yes, there has, thankfully. after all the backlash over GPT-4o and OAI's negligence companies were pressured into punishing sycophantic tendencies during the training process eqbench.com/index.html Claude models seem least sycophantic/dangerous according to independent testing

Show Post

The tech still works though. Pretending that it doesn't or that you can avoid using it to make it go away is a fool's errand. We need to organize to remove controlling power of AI from private hands, end the massive datacenter buildouts, and actually hold the misanthropic SV and VC execs accountable

Show Post