Sholto Douglas

Is that the same kind of thing that's happening to Shai GBT when he gets RLA?

I don't know.

A whole cluster of questions that can answer them and whatever.

So you, GBD7, I don't know, pulls the same thing, and then you figure out what were the causally irrelevant

What are your unknown unknowns for superhuman models?

in terms of this kind of thing where like, I don't know how are the labels that are going to be given things on which we can determine these are like this, this thing is cool.

This thing is a pay-per-click maximizer or whatever.

I mean, we'll see.

That makes me optimistic.

Do you worry about alignment succeeding too hard?

So if I think about, I would not want...

either companies or governments, whoever ends up in charge of these AI systems to have the level of fine-grained control that if your agenda succeeds, we would have over AIs, both for the ickiness of having this level of control over an autonomous mind.

And second, just like, I don't fucking trust

I don't fucking trust these guys.

You know, I don't I'm just uncomfortable with like, the loyalty feature is turned up and like, you know what I mean?

And yeah, like how much word you have about having too much control over over the eyes and specifically not you but like whoever ends up with in charge of the AI systems just being able to lock in whatever they want.

Sure.

Or alternatively, like, don't deploy when you're not sure, which would also be bad because then we just never catch it.

Right.

Yeah, exactly.