Why did China's Deepseek send shockwaves through the global tech industry?

CGTN Europe

CGTN Europe interviewed Neil Lawrence, DeepMind Professor of Machine Learning at the University of Cambridge

Transcript

00:00Neil Lawrence is DeepMind Professor of Machine Learning at the University of Cambridge.

00:05Thanks for coming on Global Business Europe.

00:07We've got lots to talk about, but let's start with this.

00:09Why did China's AI startup DeepSeek send such shockwaves through global tech?

00:16It's an interesting question.

00:18And my main answer was because tech has swallowed its own hype in terms of what I think of as AGI vaporware.

00:25And the moat around this AGI vaporware was enormous investment in compute.

00:30And what DeepSeek showed is that through agility and clever engineering and good organization,

00:36you could compete with the very best without spending enormous amounts of money.

00:41Goldman Sachs expects Chinese stocks to attract $200 billion and rise by as much as 19% this year

00:49because of the country's adoption of AI technologies, including DeepSeek.

00:54Where is China in the global AI race?

00:59So I always find this term funny race.

01:02What are we actually racing for?

01:04And, you know, this is part of the AGI vaporware sort of mantra that we're somehow racing to achieve this state of AGI.

01:13Look, each country has to decide where it wants to be.

01:16Now, undoubtedly, China is a leading force in this technology, but different countries are doing different things.

01:23Singapore is one of the leading countries in deploying the technology.

01:26So I think there's a real danger across the world in this obsession with the notion of race,

01:31rather than stepping back and saying this is a transformative technology.

01:34There's no such thing as artificial general intelligence, but this is a transformative technology,

01:39and it can take our societies to some extraordinarily good places or some extraordinarily bad places.

01:44And by not focusing on that question, we're in danger of the race leading to somewhere that none of us want to be.

01:51But do you think that DeepSeek maybe opened things up a little bit?

01:54And I wonder which other countries you expect might now dip their toe or enter the AI world with gusto?

02:01No, I think it's been a tremendous restorative to what was an appallingly polluted debate,

02:07that a smaller company has been able to show such performance.

02:10It's given an enormous amount of encouragement to companies in Europe or even smaller companies in the U.S.

02:16that they can also compete against this narrative, which is basically a Silicon Valley, Wall Street combined narrative,

02:22that you have to be enormous to succeed.

02:24So I think it's tremendously exciting.

02:26Now, I think where the next one comes from, well, I hope they're coming from a diversity of places.

02:31We've seen France has made great investments.

02:34Germany also has a lot of startups.

02:36And I'd love something to come from the U.K.

02:39What about Africa?

02:40What about India?

02:43So that's a great question.

02:44And actually, I've spent the last 10 years working with colleagues in Africa, an organization called Data Science Africa,

02:50which is building technologies developed on the continent.

02:54Now, there's lots of diversity of applications that can be used there.

02:58But what we really want is a world where people who have the problems can solve their own problems,

03:02rather than the world we're being sold on, which is one where a few companies deploy these solutions for the benefit of everyone.

03:09That has not worked.

03:10We are already in a difficult position with a concentration of power, with very few digital companies.

03:16And what this technology does offer is the possibility that everyone can interact with their machines in the ways that in the past,

03:23only software engineers were able to do.

03:25OK, so you're saying this could make it much more universal.

03:28But then I'm thinking that we're going to have to have some global AI governance.

03:31We're going to have to have some regulation.

03:33What can we expect?

03:35Yeah, I think that's the really tough bit.

03:38We've seen this enormous swing from the summits to one with this sort of almost bizarre,

03:43paranoid creation of what I think of as sort of the AI bogeyman at the UK AI Safety Summit,

03:49to sort of enormous AI boosterism in France.

03:52And perhaps that's understandable.

03:54Perhaps as politicians get to grips with the technology that they themselves don't understand,

03:58they're going to swing between these two extremes.

04:00But what I hope is more rapidly we can get to an understanding of, look, this technology is transformative.

04:05It has great potential.

04:06And there are great pitfalls associated with it.

04:09But unfortunately, a lot of those debates are not going to happen under the glare of an international spotlight.

04:14They're going to happen in the background.

04:16And one of the things I'm nervous about is more division within the research community,

04:20less access to these technologies for universities and other civil society groups,

04:24because it's those groups that will develop these technologies in the beneficial directions.

04:29Neil Lawrence, the University of Cambridge.

04:31Thank you very much.

Category

Transcript

Recommended