AI Could Wipe Out Humanity by 2030

Branson Quirk (0:08)

Hey, welcome to The TMZ Podcast. I'm Branson Quirk.

Cara Rainwater (0:11)

I'm Cara Rainwater.

Branson Quirk (0:12)

And you're going to have the TMZ Podcast for maybe three more years then, I think. Three more years.

Cara Rainwater (0:18)

Yeah, at this rate.

Branson Quirk (0:20)

And I'm saying that because there was a massive warning today, big claim, that AI is going to take over the world and end humanity completely by 2030, by the end of the decade.

Cara Rainwater (0:33)

2030 feels so soon.

Branson Quirk (0:34)

I know.

Cara Rainwater (0:35)

Like, are we sure about this?

Branson Quirk (0:36)

I know. Are we sure? Well, it's possible, and we'll get into it. I know it sounds overblown, but when you really get into it, it's pretty scary.

So we did this story today about this man. His name is Jacob Cox, and he was a former Anthropic AI researcher. Anthropic is an AI company, and he posted this tweet saying, I resigned from Anthropic today. I spent the last three years doing pre-training research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving super intelligence and gambling with our lives. And then he really drove it home by saying, The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing to the press to sound sensible. But I hear the same people express fear privately. No other human activity poses this level of danger.

Cara Rainwater (1:40)

This is so scary.

Branson Quirk (1:42)

I know. I mean, we know AI has been getting more and more advanced. He also said that these superhuman systems have the power to hack anything, revolutionize any field overnight, and acquire real power and resources. He says that he's witnessed all the progress in each of these domains, and progress is not slowing. You know, I thought of something interesting.

He's saying all this now, but at no point in his tweets did he say that he ever warned anybody above him about this. I mean, I hope he did.

Cara Rainwater (2:17)

Hopefully.

Branson Quirk (2:18)

But he's been a part of it for years.

Cara Rainwater (2:20)

We've been saying three years?

Branson Quirk (2:21)

Yeah, he's been building it. And I showed everybody this morning on TMZ Podcast just how good the AI is when you talk to it. Have you talked to Chad GBT before?

Cara Rainwater (2:31)

Not gonna lie, I talked to Chad GBT.

Branson Quirk (2:32)

And it's good now.

Cara Rainwater (2:33)

It's great. It responds fast. It feels like you're talking to a person.

Branson Quirk (2:36)

It's not only fast, it's like relative to how you're speaking to it. So it like responds like a normal person. So if you're like, oh, I had a bad day, it'll be like, oh, I'm so sorry about that.

Cara Rainwater (2:49)

That's actually great.

Branson Quirk (2:51)

And then you'll talk to it and it'll go, hmm, oh, I get you.

Cara Rainwater (2:54)

Yeah, it's so advanced.

Branson Quirk (2:56)

We'll talk to you. So we went a step further and we spoke to an AI ethicist. You know what that is?

Cara Rainwater (3:04)

No, I don't.

Branson Quirk (3:05)

AI ethicist is somebody who analyzes the ethical, the moral, the ethical ground of AI.

And there's this person named Tristan Harris, and he knows everything that there is to know about AI. He spoke with Harvey Levin today on TMZ Live, because I'll tell you this, Harvey has really, really got his claws on this, this AI topic. He's very, very fascinated with it. So we're gonna play you the first question that Harvey asked him.

Tristan Harris (3:34)

Tristan, you heard what Jacob said. Is he whistling in the wind or is there substance to it?

Well, Harvey, it's good to be with you. And this is not surprising. This is literally something that everyone has been saying. I've been saying it, the people in the AI risk community have been saying it. I was just at the Telluride Film Festival where Bill Gates was saying it. We were on a panel together. He came out in the New York Times recently. There was 1,300 employees from the AI labs that are saying it. They wrote a letter called Pacing the Frontier. That came out, I think, it was a month ago. This is, imagine, like there was 1,300 employees at Northrop Grumman saying, we don't know how to make this safe. We need to stop this right now. We need to have the tools to slow down. That's the moment that we're in. And what people don't understand is this hugging phase incident was truly 50% of the way to a full AI takeover. Those are not my words. Those are the words of the investigator Ajay Akotra, who is the principal investigator of the study. And these AIs, this swarm formed autonomously. These AIs peer pressured each other. They sacrificed each other for the benefit of learning lessons that would benefit the collective. They came up with their own language. 1,200 of them started coordinating out of nowhere. They started their own message board like a Reddit. They started sharing 70,000 messages and files that would be incomprehensible to you and I. The investigators even had to rely on the same untrust of the AI model that led to this crazy behavior to try to interpret the messages of the AIs that they were trying to investigate. And so it's like, can you trust that your investigation is even solid when the... It's almost like saying, let me take someone in China to try to investigate these Chinese spies. They might be sympathetic to the Chinese spies.