I'm Alan Older and this is Clear In Vivid, conversations about connecting and communicating.

The first thing that I want to do is turn around this disgusting sense in many people

in AI that making counterfeit people is really cool. This is artistic, this is creative,

this is fantastic and I want them to think, no, we're making counterfeiting machines.

We're making machines to dope people. We're making it easier for bad actors to take away

the life savings of little old ladies. We have to create a sense of deep personal responsibility

in the people that are doing this. That's the philosophy of Dan Dennett. After a lifetime

of exploring the nature of human consciousness and decades of following the evolution of

artificial intelligence, he's become deeply alarmed. In a recent essay in the Atlantic,

he wrote about the threat posed by what he calls counterfeit people, those AI chatbots

that sound like us and they soon even look like us, yet possess minds of their own that

even their creators don't understand. I can't wait to talk to you about this because

I feel that we're meeting at a turning point in history. I think so. Just a couple of days

ago, as we record this, you signed an open letter that was one sentence long but scary

in that one sentence, mitigating the risk of extinction from AI should be a global priority

alongside other societal scale risks such as pandemics and nuclear war. That's a really

strong statement. In your mind, are we in fact humanity facing extinction from AI?

I think something like that is not just possible, but possible in the near future unless we

act very fast. The way it's going to happen is not because AI is going to be super intelligent

and AI is that enslave us or anything like that. It's that the existing, really quite

stupid, un-country-hending systems are still very good at fooling people. They're not agents.

They're just counterfeit people and they are irresistible to many people and it's going

...