Rick Jordan

And that's what we're headed toward.

Yeah, he was.

He was the only one that actually obtained sentience.

Yeah.

That was the difference.

Oh, so he was like a prototype model.

Yeah, exactly.

He was quote unquote alive.

At the same time, the CEO recently said, because before, what, six months ago, he was like, oh, there's an 80% chance that it's self-aware.

And it was just two weeks ago, he said, I think it's about 20% now, too.

So they're starting to understand it a little bit more.

But then there's other things, too, like you're saying, that they don't understand.

So before, ChatGPT is an example.

Before it spits out a response...

Because everybody knows there's guardrails.

I mean, that's the big thing with the Department of Defense right now is they don't want the guardrails with Anthropic.

But the guardrails with ChatGPT, an example, they have agents that sit there that just act as filters.

They have specific roles.

So before there's a response that's given, it's analyzed by these other AI agents to police the original response.

And apparently some of these original responses are just super wacky.