Skip to main content

Common Sense, The Turing Test and the Quest for Real AI - Hector Levesque *****

It was fascinating to read this book immediately after Ed Finn's What Algorithms Want. They are both by academics on aspects of artificial intelligence (AI) - but where reading Finn's book is like wading through intellectual treacle, this is a delight. It is short, to the point, beautifully clear and provides just as much in the way of insights without any of the mental anguish.

The topic here is the nature of artificial intelligence, why the current dominant approach of adaptive machine learning can never deliver true AI and what the potential consequences are of thinking that learning from big data is sufficient to truly act in a smart fashion.

As Hector Levesque points out, machine learning is great at handling everyday non-exceptional circumstances - but falls down horribly when having to deal with the 'long tail', where there won't be much past data to learn from. For example (my examples, not his), a self-driving car might cope wonderfully with typical traffic and roads, but get into a serious mess if a deer tries to cross the motorway in front of it, or should the car encounter Swindon's Magic Roundabout.

There is so much here to love. Although the book is compact (and rather expensive for its size), each chapter delivers excellent considerations. Apart from the different kinds of AI (I love that knowledge-based AI has the acronym of GOFAI for 'good old-fashioned AI'), this takes us into considerations of how the brain works, the difference between real and fake intelligence, learning and experience, symbols and symbol processing and far more. Just to give one small example of something that intrigued me, Levesque gives the example of a very simple computer program that generates quite a complex outcome. He then envisages taking the kind of approaches we use to try to understand human intelligence - both psychological and physiological - showing how doing the same thing with this far simpler computer equivalent would fail to uncover what was happening behind the outputs.

For too long, those of us who take an interest in AI have been told that the 'old-fashioned' knowledge-based approach was a dead end, while the modern adaptive machine learning approach, which is the way that, for instance, programs like Siri and Alexa appear to understand English, is the way forward. But as the self-driving car example showed above, anything providing true AI has to be reliable and predictable to be able to cope with odd and relatively unlikely circumstances - because while any individual unlikely occurrence will probably never happen, the chances are that something unlikely will come along. And when it does, it takes knowledge to select the most appropriate action.

Highly recommended.


Review by Brian Clegg


Popular posts from this blog

The Feed (SF) - Nick Clark Windo ****

Ever since The War of the Worlds, the post-apocalyptic disaster novel has been a firm fixture in the Science Fiction universe. What's more, such books are often among the few SF titles that are shown any interest by the literati, probably because many future disaster novels feature very little science. With a few exceptions, though (I'm thinking, for instance, The Chrysalids) they can make for pretty miserable reading unless you enjoy a diet of page after page of literary agonising.

The Feed is a real mixture. Large chunks of it are exactly that - page after page of self-examining misery with an occasional bit of action thrown in. But, there are parts where the writing really comes alive and shows its quality. This happens when we get the references back to pre-disaster, when we discover the Feed, which takes The Circle's premise to a whole new level with a mega-connected society where all human interaction is through directly-wired connections… until the whole thing fails …

Everything You Know About Space Is Wrong - Matt Brown ****

What we have here is a feast of assertions some people make about space that are satisfyingly incorrect, with pithy, entertaining explanations of what the true picture is. Matt Brown admits in his introduction that a lot of these incorrect facts are nitpicking - more on that in a moment - but it doesn't stop them being delightful. I particularly enjoyed the ones about animals in space and about the Moon.

Along the way, we take in space exploration, the Earth's place in space, the Moon, the solar system, the universe and a collection of random oddities, such as the fact that Mozart didn't write Twinkle, Twinkle Little Star. Sometimes the wrongness comes from a frequent misunderstanding. So, for example, Brown corrects the idea that Copernicus was the first to say that the Earth moves around the Sun. Sometimes there's some very careful wording. This is used when Brown challenges the idea that the Russian dog Laika was the first animal in space. What we discover is that, i…

Dark Matter and the Dinosaurs - Lisa Randall ****

I did my PhD in galactic dynamics - which is an awkward subject when people want to know what its relevance to the 'real world' is. So I was excited when Clube and Napier's book The Cosmic Serpent came out, around the same time, because it provided me with a ready-made answer. It argued that the comets which occasionally crash into Earth with disastrous results - such as the extinction of the dinosaurs - are perturbed from their normal orbits by interactions with the large-scale structure of the galaxy.

I was reminded of this idea a few years ago when there was a flurry of media interest in Lisa Randall's "dark matter and the dinosaurs" conjecture. I was sufficiently enthusiastic about it to write an article on the subject for Fortean Times - though my enthusiasm didn't quite extend to purchasing her hardback book at the time. However, now that it's out in paperback I've remedied the situation - and I'm glad I did.

Dark matter is believed to exi…