This is the Guardian.

Today, a warning from the Godfather of AI over the profound dangers of the technology he helped to invent.

Are you okay with me recording this?

Development in AI are moving so quickly.

And there's this assumption that the scientists who have been working on this stuff for decades.

They know what they're doing.

Last week, the Guardian's UK Technology Editor, Alex Hurn, tracked down one of those scientists.

In fact, arguably one of the people most responsible for getting us to this point.

They met not in some futuristic office in Silicon Valley, but in a house in the suburbs of North London.

When the podcast uses it.

When I arrived in his house, he had me sit down at his sister's kitchen table.

And then we had an hour-long conversation which honestly felt most like a university tutorial.

I have a lot of the same concerns as lots of other people.

I just give you a list of all those sort of what I consider the standard concerns.

So, Jeffrey Hinton's long been known as one of the three Godfather's of AI.

It's a grand years title, but it's earned.

He won the ACM Cheering Medal, the Computer Sciences Equivalent to the Nobel Prize in 2018.

And it was for work on deep learning.

The fundamental approach to AI that you'll also have heard referred to as neural networking,

which underpins almost everything that has laid at the heart of the revolution we've had in the last couple of years.

From chat GPT, to Dali, to Google Translate.

Beyond being one of the centerpieces of the recent history of AI, though,

...