Annie Jacobsen

Since AI is based on machine learning, from my understanding, it has to get its information from somewhere to then build new information like the chats.

Right.

Okay.

So if you follow that logic, then wouldn't it follow that...

What it is learning is all of humans bad behavior.

No, I don't think so.

It'll see actions and consequences over hundreds of all of recorded history.

And then possible future consequences.

I haven't read Ray Kurzweil's new book yet, but I wonder if I might disagree with that, only that I'm thinking that everything you're saying would be true once singularity, or for laymen, once AI, once a machine figures out how to think for itself, makes that leap, which is almost like an unknowable...

presently incomprehensible to me jump where it can think for itself.

It's almost so hard to even comprehend what that is.

But based on our languages.

So... For now.

For now, right.

But that's what I'm saying is where they can suddenly... Where they experience that moment in time that you and I were talking about earlier where man went from pointing...

to suddenly using a symbol and realizing in his own brain, wait a minute, I can make this symbol represent

a sound, and then I will put it together to make it a language that only my tribe can understand.

That to me is like a giant leap in humanness that I think about often and can't really, I can't understand how that happened other than, wow, how did that happen?

That seems to me that that has to happen to the AI before it can really think.

Otherwise, it's just basing its thinking on recorded history.