How should Artificial Intelligence develop?

CGTN Europe

At the Artificial Intelligence Action summit in Paris, a key talking point has been the development and how to fund and prepare.   There have also been discussions about moving forward, building larger data centers and providing enough energy for them.   Among several side bar events was China’s AI Development and Safety gathering.   Iolo ap Dafydd reports from Paris.

Transcript

00:00Paris has been full of tech company executives and world leaders grappling with how artificial

00:07intelligence should develop.

00:09Over the past few days, all kinds of ideas and initiatives have been discussed.

00:14Professor Song Yi has returned to Paris to discuss the rapid development on AI by some

00:19tech corporations.

00:21Many of the AI companies now, they choose not to talk about AI ethics and safety, but

00:27I think this is completely wrong.

00:30Actually the relationship between AI development and safety is not that they drag each other,

00:36they will have negative influence to each other, no.

00:39Actually you can build AI which is very powerful from the development point of view, but still

00:45you keep it really safe.

00:47So this is scientifically solid.

00:50The Paris Summit's declaration seemed to antagonize both the US government with references to

00:55diversity, gender and disinformation, as well as the UK government, by ignoring the

01:00scientific and political consensus around risks from smarter-than-human AI systems that

01:06was agreed at Bletchley Park in late 2023.

01:10One argument is to allow companies to innovate.

01:14Now today we're in the absurd situation where there are more safety standards on a San Francisco-based

01:18company that wants to sell sandwiches than the ones that sell superintelligence.

01:24It's completely nuts.

01:25As soon as you have the safety standards there, companies will innovate to meet them and give

01:30us powerful AI that's actually safe tools that work for us, rather than completely useless

01:37superintelligent things that just make us obsolete and take over.

01:42As important as the main summit were a myriad of other events organized throughout central

01:47Paris, debates about AI, after speeches, in private meetings and panel discussions

01:54on the innovation and how to develop it safely for the benefit of everybody.

01:59At this panel, organized by China's AI Safety and Development Association, many experts

02:05seem to agree there should be international cooperation on artificial intelligence and

02:11transparency too.

02:13The solution is in our hands.

02:15We can decide which direction we want to go, how far can we go, and how quickly should

02:20we go.

02:21So I think that's where we feel that AI governance, particularly at the global level, is critical.

02:29President Macron, in his address at the end of the first day of the summit, injected some

02:33humor into the French government's intention to invest US$112 billion in new data centers

02:40and power for AI technology in France.

02:43I have a good friend in the other part of the ocean saying, drill, baby, drill.

02:49Here, there is no need to drill.

02:52It's just plug, baby, plug.

02:55Electricity is available.

02:57You can plug it.

02:59It's ready.

03:00The balance and the challenges of how to regulate and by how much, while allowing tech companies

03:06to innovate, will continue long after this AI Action Summit.

03:11Yolanda David, CGTN, Paris.

Category

Transcript

Recommended