'We can fix the internet' says Tim Berners-Lee—just don't ask what he thinks about AI

Rejoice all ye of fading faith in the internet. No less an authority on the subject than Tim Berners-Lee, credited by many as it inventor, thinks it’s not to late for a fix. But you might like to skip his prognostications on AI. They’re a little less optimistic.

In an interview with the Guardian as part of his ongoing book tour, Berners-Lee strikes a pretty positive tone about the internet, despite the fact that he says it is now “optimised for nastiness.”

“We can fix the internet,” he says, “it’s not too late.” The problem, of course, is commercialisation. “The Americans were very keen about commercialising the internet, crossing the boundary from being an academic thing to being a commercial thing,” Berners-Lee explains.

Tim Berners-Lee's computer

It all started here, and it’s kinda gotten out of control… (Image credit: Coolcaesar at the English-language Wikipedia)

“I’m interested to see how it goes down in Australia, because there are people in the UK who are proposing it in a similar way, and others may follow on. The first question is whether kids should be using those particular social media sites. I think you have to recognise that things like messaging services are useful,” he says.

But he is already in favour of devices for kids that limit access to potentially harmful sites and social media platforms.

Of course, when it comes to AI, the irony for the likes of Berners-Lee is that the current boom in generative models is arguably only possible because of the internet. Without the internet, it wouldn’t be practical to acquire the huge quantities of training data required.

Speaking of irony, contrary to his decentralising approach to the internet, he wants to bring AI development under much closer control. “I would like to see a Cern for AI, where all the top scientists come together and see whether they can make a super intelligence. And, if they can, they contain it into a system where it can’t just go out and persuade people to let it run the world,” he explains.

Without that, he says he can’t, “see a way that we can get to a point where the scientific community gets to look at the AI and to decide whether it is safe or not.” As things stand with AI, “the horse is bolting.”

All of which means that even if the internet can be saved, it could all be for nought if AI gets us anyway. A cheery thought, eh?