Should we trust tech titans to control AI?

Alex Ritson (0:00)

This is the Global News Podcast from the BBC World Service.

I'm Alex Ritson, and in the early hours of Wednesday, the 16th of September, these are our main stories. Two conferences on AI revealed different views on how to control the technology. The US Defense Department confirms it's facing a shortage of precision weapons because of the war with Iran. And the authorities in Norway searched the headquarters of the telecom operator Telenor for helping the military government in Myanmar. Also in this podcast, the tech giant Metta takes action after a BBC investigation found adverts for child abuse material on Instagram.

Divya Arya (0:45)

Child rights bodies and government agencies have now been demanding that Metta needs to make its guardrails better and stop the material even before it's published on their platform.

Alex Ritson (0:57)

And the support acts for pop star Ed Sheeran walk out of his US tour in Aarau over pro-Palestinian comments.

Shirley Halperin (1:04)

He rightfully said he's appalled by, you know, the conflict that's happening in the Middle East. At the same time, he's very apolitical for a reason.

Alex Ritson (1:17)

We start with a debate that's dominated the headlines over the last week. How safe is artificial intelligence?

Two camps appear to have emerged offering conflicting views on how to deal with the fast developing technology. On one hand, there's a push for more regulations voiced at a conference in Washington DC on Tuesday by the US Senator Bernie Sanders. He accused his congressional colleagues of being too slow to act on the risks posed by AI.

Bernie Sanders (1:46)

Despite living in a so-called democracy, the public has had virtually no input into the AI revolution that is transforming the world in which they live. Congress under both Democratic and Republican control have been asleep at the wheel, and we now have a president whose ignorance regarding this issue is truly embarrassing.

Alex Ritson (2:19)

Just a few hours after that gathering, tech leaders and bosses were speaking at another AI conference in San Francisco.

Some executives have themselves called for tougher regulations, including the boss of Anthropic. But others are hoping their businesses can carry on with few constraints. One of those is Yen-Sung Huang, the chief executive of chipmaker Nvidia. He told the conference that the market can regulate itself.

Jensen Huang (2:45)

If you build a product or a service and you're not confident in its functionality, capability or safety, then don't release it.

And so that's a very obvious thing to do. You pace yourself until you are confident you're releasing something that the market would appreciate. The market forces are already there. We don't need any new laws. We don't need any new regulations.

Alex Ritson (3:06)

Sam Altman, co-founder of Open AI, has also been speaking, admitting that the speed at which AI is developing is scary, but also urging the world to trust the AI industry to do the right thing. But can we? Our North America technology correspondent, Lily Jamali, told me what vision both camps have been laying out.

Lily Jamali (3:27)

They're pretty much staking the same positions that they already have. You have the very strange bedfellows in a progressive senator, an independent but a very liberal Senator, Bernie Sanders, coupled with the man who helped get Donald Trump elected in 2016, Steve Bannon, who is very conservative, joining forces in DC at that one conference saying, we need regulation, these billionaires can't be trusted. Then on the other hand, here in California, I'm in Los Angeles, but up in San Francisco, the Dreamforce Conference held by the company Salesforce, you heard AI leaders, people with the most skin in the game, pretty much singing from the same songbook saying, we've got this, you can trust us.

Alex Ritson (4:12)

It was a shock earlier in the week to hear these predictions from your senior people in the industry that there is a good chance, they say, of AI agents killing all humans within 10 years. How have people inside and outside the industry been digesting that news over the last week?

Lily Jamali (4:33)

Well, Jacob Cox and the AI safety researcher who abruptly resigned from Anthropic, one of the leaders in the AI space right now, his resignation, I think, was really galvanizing in terms of starting a public dialogue that really has not existed in this way. Yes, we talk about AI, we talk about data centers, but I think the concept of human extinction at the hands of this technology is one that is only now starting to trickle into a more mainstream culture.

I was talking to a CEO at a different conference last week who told me that is very much the culture at that company, at Anthropic. There is, at least from where others sit, a Doomer culture. They have marketed themselves as being safety first, and so the cynics will say, well, that's because that's a way of selling their product. But I will say, Dario Amode, who started that company, he left Open AI because he also felt that safety wasn't being centred in the development of the technology there, the maker of ChatGPT. So he goes off and co-founds this other company. And so I think that that view is maybe a little more common in Silicon Valley than the outside world might have known.