Jonathan Haidt

So I think it's chutzpah of Meta to say, we caused this problem and now we have a tech solution for it.

Plus, there are also a document just came out last week about Meta's internal ethical standards for AI.

Their chief ethicist signed off on this, that the platforms can have sensual conversations with children, that it can talk about sex with them.

There are some limits that you can get around, but it can show them violence, it can talk sex with them.

So these are companies that have shown over and over again, they cannot be trusted,

Plus the chatbots aren't really programmed.

We don't know why they're doing what they do.

They're equivalent of neural networks that get tuned up and they all have unpredictable properties.

So even ChatGPT, which has some safeguards, you know, there's an article in the Times, a woman found out how to have a deeply erotic affair with ChatGPT in terms of, you know, sexual fantasies and rape fantasies and things like that.

So the idea that, you know, here I am working so hard with my team

to get across the idea, our children should not be on smartphones and social media.

They should not be talking to strangers.

They need a normal human childhood.

And we're making a lot of progress.

Then all of a sudden, in the last couple of months, everything is flooded with chatbot toys.

Mattel is teaming up with OpenAI so that their toys will have ChatGPT in them.

Your daughter can talk to Barbie.

Barbie will literally be her best friend.

And then good luck taking away her Barbie because it is now literally her best friend.

So I think we're gonna have a lot of success, but only in certain places.