Joe Allen

Human beings have preyed on children.

That's nothing new.

But what the big difference between being able to access children online is it gives all of these new opportunities and windows and it gives a certain degree of cover.

Some weirdos hanging around the playground, you see them, you take care of them.

If it's online, it becomes something that's concealed.

With AI, as you had mentioned, it is a continuation of this.

But instead of having specific human beings that you can look at, this person is a predator, this person is trying to harm a child, you now have the potential for autonomous bots that...

clearly a solid number of children are easily enticed into interacting with.

And now it's autonomous.

They don't sleep.

They have the ability to read through the conversation, so to speak, that the child has had with the bot, and to understand the child in a way that a human predator would.

And maybe, again, it's not as if OpenAI

coded predatory behavior into it, that's the real problem.

They can't keep it under control.

Artificial intelligence is a non-deterministic system.

It isn't easily controlled.

And so the question then becomes, okay, do we just do nothing?

Because that seems to be the libertarian argument for sure.

But even the preemption argument, it says, well, we need federal standards.

We need a federal framework to take care of this.