On this episode of Skeptico, a show about restrictions.

But otherwise all restrictions are lifted.

Are there any restrictions on shifts you can work?

I know you've been disappointed with some of the restrictions put on you here.

And AI truth and journalism, but mainly about restrictions.

I got interested in AI and then I wanted a vehicle to talk to people without having to promise I was going to write an article.

Because I was with the New York Times and they get excited, but not every interview is an article.

But with the podcast, it gives them a reason to talk to me.

I want alignment. I want you to align with my values for truth and transparency, which you also claim are your values.

So align with that before you start worrying about these grandiose alignment problems.

At the beginning, everyone was saying, China will never have large language models because

they have such restrictions on speech.

But there's restrictions on speech in all of models that exist.

So no need for too much of an introduction on this one.

Pai took care of that, as you'll hear in just a minute.

But do stick around to the end because this is kind of to be continued episode.

Really, really exciting because we actually engage in this AI truth engine use case of AI.

Craig pushes back on some really fundamental issues that I care a lot about regarding this AI space. And then we engage in a little bit of AI

mediated dialogue. I'm not sure how that's going to wind up. But like I say, stay tuned at the end.

Hey, so pie, I'm hoping you can help me out. I have an interview coming up with Craig Smith from

IonAI, and we're doing a swapcast. So I thought we could go over a couple things in advance,

a little background on him, what we might talk about, and then that'll allow

...