Demis Hassabis - the artificial intelligence guru worth £400m to Google

Demis Hassabis Image copyright DeepMind

Chess-playing Londoner reveals his plans for computers "that think"

Two weeks ago, I had the pleasure of spending a few hours in the company of a man who, though not at the moment a household name, is likely to be one of the most important figures in the way the world of computer technology develops in the next 20 years.

My film of the interview is being broadcast tonight on Newsnight (though not the bit, thankfully, where he humiliates me at chess).

Demis Hassabis is the founder and chief executive of Deep Mind, the artificial intelligence business bought by Google for £400m at the start of the year.

He is also a chess master (so maybe my crashing loss to him was not such a humiliation), the developer behind some of the biggest games of the 1990s including Theme Park, and the Usain Bolt of the annual Mind Sports Olympiad, an event where now apparently even his children are starting to impress.

His day job is now looking at how artificial intelligence (AI in shorthand) will develop our understanding of data and the way computers make sense of the world around us.

Demis - it would seem odd to call him Mr Hassabis given that he wears t-shirts and looks a bit like your classic Silicon Valley entrepreneur explains AI by way of Deep Blue.

You'll remember, that was the computer that famously beat chess master Garry Kasparov in 1997.

But as Demis points out, Deep Blue was very good at chess because it was programmed to be. It couldn't, however, play noughts and crosses.

Artificial intelligence is about changing that.

It is about enabling computers to "learn" and understand vast swathes of data for example on climate change, health or economics beyond the capacity of the unaided human mind. And then to be able to apply what it has learnt to other fields.

The development of such "thinking computers" is of course fraught with difficulty. One only has to think of the possible military applications to get a sense of that.

And Terminator is a film we all remember well.

Demis reveals to me that he is well aware of the concerns. So aware, that Google is setting up an ethics committee to look at the possible implications of artificial intelligence.

In a similar way to health research, where this work goes will become of vital importance in the next 20 years.

Google has already said that it will not engage in military research. Demis said that this commitment was one of the reasons he agreed to sell Deep Mind to them.

Regulatory oversight by governments will also surely be necessary. This cannot simply be left to the companies themselves to police what they are doing.

At some point, this will be a major public debate. For so important an issue it is all rather sotto voce at the moment. It is time governments and other interested bodies started focusing.

Related Topics