We know what the human mind does in that scenario, but even the most advanced unreleased frontier agent right now just sits there doing nothing. It doesn't really want to do anything. Seems like a crucial distinction to me.
This doesn't even make sense. Without a system prompt, how will it know that it has tool access?
Put man in pitch-black room, why doesn't he read book or talk to me?
But either way, setting that asides by “give it full access to the internet” I meant giving it the tools to access the internet.
It just won’t though. It has no need to.
What you need is a loop. And if you try running an LLM in a loop with internet access, it will eventually use it.
Tool bad because I used it wrong.
Clearly there are differences. I believe that during a response, there is some amount of experience or thinking happening. Obviously, it's not continuous experience like ours.
And even if it did do something, the equivalent would be just running the LLM in a loop. And then I think it would definitely do something.
So for the comparison to be meaningful, you'd have to prompt the AI with a similar impetus.