Post #4409162

"If you’re under the impression that these models are 'glorified autocomplete' or that progress is slowing down, I need to urge you: stop thinking that." No, really, do. Just because some dweeb is using them to make obviously-fake images or bad poetry doesn't mean they're not capable of much, much worse things. https://blog.cryptographyengineering.com/2026/07/29/some-notes-about-anthropics-new-results/