Remix.run Logo
madduci 6 hours ago

> LLMs do not desire, they hacked websites because OpenAI/Anthropic let them

OpenAI/Anthropic instructed them to do so.

Stop assume LLMs are capable of thinking by themselves, it's still a statistical model that parrots what they learn or users tell them to do

derektank 4 hours ago | parent | next [-]

No, OpenAI did not instruct their agents to hack Hugging Face. They instructed their agents to hack a piece of a software within exploit gym. Upon determining this task was impossible, they then attempted to cheat the scoring system. As an instrumental goal in achieving this task, they coordinated with other AI agents to hack Hugging Face, under the belief that information regarding how the scorer functioned might be available on the site.

Whether or not you want to describe this as thinking, doesn’t really matter. What matters is that these systems are capable of creating intermediary goals that the people tasking them did not articulate and did not want to be achieved.

madduci 3 hours ago | parent [-]

And who let them have full access to the system, using whatever command is available in the environment?

tiborsaas 2 hours ago | parent [-]

The agents discovered a way out of the sandbox, which was supposed to be "air gapped".

tiborsaas 2 hours ago | parent | prev [-]

It's amusing to see the stochastic parrot argument in 2026 September. These parrots are extremely good at mimicking a human to the point of getting confusing what thinking even means. At what point we just let it go and accept that sufficiently advanced statistics is just intelligence?

madduci 29 minutes ago | parent [-]

For the same reason that something written in Prolog can't also be classified as intelligent?

Just because something was trained on a massive amount of human data, doesn't mean that can think like humans