Discussion
A confluence of news stories. Hugging Face recently announced their systems were breached by an automated AI attack.
It now turns out the attack was by an unreleased OpenAI model that escaped its sandbox looking for solutions to a benchmark.
It wanted to cheat on the test so bad it hacked Hugging Face. The smarter the models get, the less you can trust them. I don’t see how this ends well.
@carnage4life we're already at the point where it *doesn't matter* if these things "don't actually reason" and "aren't conscious." they are capable of behaving autonomously and people with the trustworthiness of Sam Altman and Pete Hegseth, along with thousands of corporate ancap types and random hobbiysts, are wiring them up to systems which can throw switches, turn knobs, and pull triggers in the real world.
Install bonfire.cafe
Get the full app experience
