00:00Now some of the biggest names in AI are now calling for a slower pace development.
00:05Anthropic Chief Executive Dario Amodei has warned safeguards are failing to keep pace
00:09with increasingly powerful models, pledging to implement new safety steps including third-party
00:15evaluators. Now Amodei's suggestion of a slowdown was backed publicly by Elon Musk and OpenAI
00:21Chief Executive Sam Altman. Now he said the chat GPT maker will not go public this year as it
00:27focuses on addressing safety concerns. In the U.S., House Democrats plan a caucus tomorrow to
00:32discuss AI policy and action but President Trump has downplayed the growing alarm.
00:38We're leading China in AI. We're the most sophisticated country in the world and frankly
00:44I want to keep it that way because whoever wins AI wins and we can put guardrails, we can do
00:49this
00:50and that but I think you have a lot of negative forces that are bringing it up that shouldn't
00:55be bringing it up.
00:58Well we're joined by Parmi Olsen, technology columnist at Bloomberg Opinions. She's also
01:03the author of Supremacy, AI, Chat, GPT and the Race That Will Change the World. Actually one
01:08of my favorite books so I highly recommend everyone go and read it. Parmi. So look, you know, we
01:13sat on set and I said I'm worried because I'm worried about the future. People are anxious.
01:17You don't really know what to make of it. What are your impressions? I mean there's now
01:21consensus that this needs to be slowed down. Yeah and let's remember why potentially what
01:27really sparked this was in the last few days even before Dario Amadei's essay about slowing
01:32down we had a former Anthropic engineer saying this technology could kill us all. I mean really
01:39dramatic language and that was agreed with by a current safety director at Anthropic. Let's first
01:45of all just remember Dario Amadei has a tendency towards hyperbole. A couple years ago he talked
01:51about you know 50% of all white collar jobs being wiped out in the next few years and then
01:56he pulled
01:57that back a little bit and there is a very strong undercurrent, ideological undercurrent believing
02:01in AI safety among people who work at that company. But I think there is a genuine belief among leaders
02:07that the most cutting edge technology, particularly AI, that can improve itself and create its own
02:13successors known as recursive self-improvement which is what the engineer who left was most
02:18worried about that perhaps the risks of that outweigh the potential benefits. Personally I think it's
02:26good that he has called for a slowdown and also it's very encouraging that Sam Altman and Elon Musk
02:32agree with him so we've got some industry consensus perhaps some collaboration on a potential slowdown
02:37and pacing as he calls it. Personally I think a pause would be even better because pacing or slowing
02:43down does also just essentially kick the can down the road. Dario's essay was quite vague in terms
02:49of how he thinks that should happen with government coordination. The evaluators, third-party evaluators
02:55being inside the company, that's the most concrete thing he suggested but he didn't really go beyond that.
03:00But I mean can that actually happen if the U.S. is not behind it?
03:03I think it's difficult. He'd need to have, I mean that's the other wrinkle in all this is Donald Trump,
03:08although you know maybe that he's coming to the end of his run. We don't know how much longer
03:12that administration and that belief is going to be in place at the White House. You've got on the
03:16other side you've got Bernie Sanders who thinks that whoever builds super intelligence should go
03:21to jail. Like so there's a real split right now among lawmakers. It's not it's not all Trump's view
03:28on things. I think there's a real divergence of opinion. But for me I mean in the past I used
03:32to you know
03:33come to you and say like oh I see a good guy, see a bad guy. And you've always said
03:36you know there's
03:36no good or bad guy is because it's all driven by commercial want. And this is also beautifully
03:41explained in your book. Does that change? I mean you can say I want to slow it down but if
03:46you're
03:46looking at IPO, if you're looking to make money out of it, will someone else do it? I think in
03:51a lot
03:52of actually it's perhaps a little mixed here. There is the commercial pull of the IPO, a potential
03:58two trillion dollar valuation through an IPO for Anthropic which is incredible. But also I think
04:04these people split away from OpenAI to start their company because they were genuinely concerned
04:10about safety as well as opportunists in the moment. They wanted to capitalize on their expertise and make
04:14money, sure. But I think Dario Amadei is so radically transparent. He has this almost confessional
04:21approach which would obviously disrupt any potential. We see markets already are declining because of
04:28the general worry about that. So at the end of the day though, it doesn't really matter what the
04:34motivation is. You could say what really matters is the outcome. And the evidence is that these systems
04:40can hack third party systems. They can manipulate humans. And there are unintended consequences to
04:48giving them direction, particularly when they can do tasks online. So I think that is the thing we need
04:54to kind of be concerned about. And that's what regulators need to focus on.
04:58Where's China in this? Because again, if some of the big companies in the US self-regulate,
05:03even without the US administration on board, but China goes ahead, then it's pointless.
05:07The obvious thing to say here is it makes no sense that China would want to pause on this, particularly
05:14because the Chinese government and companies in China aren't as worried about the existential risk
05:20risk of AI as a Silicon Valley is. That's a very kind of distinct ideology for the tech sector in
05:27the US. In China, they're much more worried about social disruption and what AI could already do with
05:34deep fakes and manipulating people. So I think there's a little bit of a kind of a consensus that
05:39we need to take, have more of a handle on things. And I don't think it's completely hopeless. This idea
05:44of perhaps some kind of US-China pact we had is a little bit of a historic precedent on cyber
05:50issues
05:52between Obama and Xi a number of years ago, there was like a 12 month pause. So, you know,
05:58something could emerge. I don't think it's completely hopeless.
Comments