This post was also posted yesterday, but yeah, with ChatGPT you can execute random code. It is however in a VM of some sorts, so just trying to delete things won’t do that much.
It doesn’t kill the LLM instance, internally it just calls an API to run the generated code on a machine if you ask for it.
With chain of thought it basically asks itself to generate related sub questions and then answers for those sub questions.
Basically it’s just the same but recursive. So, like it looks like it can tell you things, it just also looks like reasoning.
Now it may well be an improvement, but it’s still basically. “I have this word, what is statistically most likely to be the next word” over and over again.