Technology and People: Deloitte Report

Lord Rees of Ludlow Excerpts

Wednesday 13th April 2016 Lords Chamber

[Lord Rees of Ludlow (CB)]

My Lords, the big economic and social question that should concern us is surely this: will robotics and AI be like earlier technologies addressed by Deloitte and create as many jobs as they destroy, or might it really be different next time? Robots have already replaced people in much of manufacturing, but in coming decades they will take over not just manual work—indeed, jobs such as plumbing and gardening will be among the hardest to automate—but routine legal and accountancy work, medical diagnostics and even surgery.

DeepMind, a small London company now hoovered up by Google, hit headlines recently because its computer beat the world champion in the Chinese game go. This was a breakthrough in so-called “generalised machine learning” because, unlike IBM’s earlier chess-playing computer, the go-playing machine was not programmed by experts. It taught itself by analysing lots of games and playing against itself.

Computers learn to identify dogs, cats and human faces by crunching through millions of images—not the way babies learn. Computers learn to translate by reading millions of pages of, for example, multilingual EU documents—they never get bored. But advances are patchy. Robots are still clumsier than a child in moving pieces on a real chess-board. They cannot tie your shoe-laces or cut your toe-nails. But sensor technology, speech recognition and so forth are advancing apace. The driverless car is closer to reality.

Incidentally, the driverless car raises questions of safety and how to cope with emergencies. For instance, if an obstruction suddenly appears on a crowded highway, can Google’s driverless car discriminate whether it is a paper bag, a dog or a child? The likely answer is that its judgment will never be perfect, but it will be better than the average driver. Machine errors will occur, but not as often as human error. But when accidents do occur they will create a legal minefield—who should be held responsible: the driver, the owner or the designer?

According to the Deloitte report, the jobs that have multiplied most since the 1990s are socially valuable but poorly paid—nursing auxiliaries, teaching assistants and care workers. This trend will continue. The money “earned” by robots could generate burgeoning wealth for an elite, but sustaining a harmonious society will require massive redistribution to guarantee everyone a “living wage”, and to expand and upgrade public service employment where the human element is crucial for our quality of life and is now undervalued—carers for young and old, custodians, gardeners in public parks and so on.

I will indulge myself briefly by looking further ahead. Today’s smartphones would have seemed magic just 20 years ago, so looking towards mid-century we must keep our minds open, or at least ajar, to what may now seem science fiction. Some AI evangelists talk about an intelligence explosion when machines will achieve human capabilities and will then go on themselves to create even more intelligent machines. Just as a nuclear explosion is easier to create than to control, maybe there is a worry that an intelligence explosion would be harder to control, even if it could be developed.

What if a machine developed a mind of its own? Would it stay docile or might it “go rogue”? Could it infiltrate the internet and the internet of things and manipulate the external world? This may seem science fiction, but some AI pundits think that the field already needs guidelines for responsible innovation, just as biotech does, to ensure that we can cope with or prevent these obvious downsides. Others regard these concerns as premature and think that it will be several decades before artificial intelligence becomes more of a worry than real stupidity.

But the disagreements are basically about timescales—the rate, not the direction, of travel. Few doubt that machines will surpass more and more of our distinctive capabilities or enhance them via cyborg technology. The jury is out on whether they will be idiots savant or develop truly versatile intelligences. Either way, coping with their societal consequences will become ever more challenging than it already is. It therefore needs to be high on our agenda. That is why I hugely welcome this debate.