The Guardian view on robots and humanity: passing Go | Editorial

Machines can’t feel or do many of the things that make us human. Sadly that doesn’t dispel the concerns about them putting people out of business

The defeat of Lee Sedol, the world’s strongest Go player, by a Google AI program, looks like another milestone towards a world where computers can do almost anything a human can. It is not. There are uncountable things that only a human can do, and that no computer seems close to. The problem is that the purely human things are not economically useful to anyone. The things that computers can be taught to do are by contrast economically fantastic. The most powerful programs in the world are no more human than a shovel is, or a nuclear submarine. They are not moral actors and they have no feelings. What they have is power, but this power is growing at a rate that should frighten us all.

It might be less frightening if computers were genuinely alien and intelligent, much as HG Wells imagined the Martians – “Minds that are to our minds as ours are to those of the beasts that perish, intellects vast and cool and unsympathetic” – but even the most powerful networks are less human than Wells’s monstrous Martians. Their power will be used to make money for the firms that finance their development, and then for others quick and clever enough to take advantage of the new world. It’s hard to imagine them being used to dent inequality of either wealth or power, globally or within countries. It is far more likely that they will increase inequality and still further hollow out the middle classes as we move towards an hourglass society in which everyone is either very rich or very poor and likely indebted.

Facial recognition software is a magical-seeming example of artificial intelligence, but its use for law enforcement and passport scanning can be put to the service of repressive regimes. Even when giant companies – giant in their profits, not in the numbers they employ – map our friendship networks, so that we can more efficiently be sold things, this is not progress.

One of the more sinister effects of the spread of more intelligent computer networks is the simultaneous spread of what might be called artificial stupidity. Given that machine learning is deployed largely to replace unskilled labour, it is most productive when labour is kept unskilled or redefined that way. IBM is testing a robot concierge in a Hilton hotel, something that is both a gimmick and a premonition. So much of the work in service industries is now simplified until it might be automated. And robots will never need pensions. The automation of employment and the inexorable rise of zero-hours contracts, confirmed in new figures on Wednesday, are both part of the same process. AI is nibbling away at skilled work, like some forms of medical diagnostics, at the same time as older doctors complain that the traditional human skills of diagnosis are falling out of medical training. The belief that everything worthwhile can be measured and then managed is far more damaging to humanity than the threat of artificial intelligence on its own, but “deep learning” techniques are increasingly deployed in the service of this belief.

But no triumph in complicated games can bring us closer to genuinely autonomous and conscious computers: the test for true sentience would be a program bewildered and frightened by the knowledge of mortality.



The GuardianTramp

Related Content

Article image
The Guardian view on artificial intelligence: look out, it’s ahead of you | Editorial
Editorial: There is a tendency to see intelligence where it does not exist. But it is just as wrong to fail to see where it is emerging


08, May, 2016 @7:41 PM

Article image
Laughing parrots, backflipping robots and saviour viruses: science stories of 2017
Leading scientists pick the dozen most significant discoveries and developments of 2017

Liz Sockett, Sophie Scott, Adam Rutherford, Jim Al-Khalili, Amy Dickman, Beverley Glover, Callum Roberts, Jan Zalasiewicz, Danielle George, Mark Miodownik, Helen Czerski and Kevin Fong

24, Dec, 2017 @9:15 AM

Article image
AlphaGo seals 4-1 victory over Go grandmaster Lee Sedol
DeepMind’s artificial intelligence astonishes fans to defeat human opponent and offers evidence computer software has mastered a major challenge

Steven Borowiec

15, Mar, 2016 @10:12 AM

Article image
Can Google’s AlphaGo really feel it in its algorithms? | John Naughton
When game-playing system AlphaGo defeated a master of the Chinese game go, its creators could not explain why. Is this a sign of intuitive AI?

John Naughton

31, Jan, 2016 @9:00 AM

Article image
Google's artificial intelligence machine to battle human champion of 'Go'
Lee Se-dol, 33, one of the world’s top players of the ancient Asian pastime, is confident he can beat Alphago. But he hasn’t seen improvements made to the system

Alex Hern

07, Mar, 2016 @2:33 PM

Article image
AlphaZero AI beats champion chess program after teaching itself in four hours
Google’s artificial intelligence sibling DeepMind repurposes Go-playing AI to conquer chess and shogi without aid of human knowledge

Samuel Gibbs

07, Dec, 2017 @12:41 PM

Article image
Google's AI machine v world champion of 'Go': everything you need to know
Alphago, a cutting-edge computer program designed by Google, will face off in the Google DeepMind challenge against Lee Sedol, the world champion of Go

Steven Borowiec in Seoul

08, Mar, 2016 @11:30 PM

Article image
Fat Cat Thursday and the changing world of work | Letters
Letters: Tim Gossling says that automation means the old capitalist model is no longer viable; Albert Beale writes that our drive for ‘efficiency’ is creating an increasingly inhumane world; plus letters from Keith Flett, John Wilson and Christine Weaser


04, Jan, 2018 @7:25 PM

Article image
Why the rise of the robots could allow humans to flourish again | Giles Fraser: Loose canon
Loose canon: Nobody’s job is safe. But a citizen’s income in a post-work world could see us avoid the Terminator scenario and return to pre-capitalist sources of value

Giles Fraser

31, Aug, 2017 @3:05 PM

Article image
The superhero of artificial intelligence: can this genius keep it in check?
With his company DeepMind, Londoner Demis Hassabis is leading Google’s project to build software more powerful than the human brain. But what will this mean for the future of humankind?

Clemency Burton-Hill

16, Feb, 2016 @8:15 AM