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.

Hardback:  

Kindle 
Review by Brian Clegg

Comments

Popular posts from this blog

Beyond Weird - Philip Ball *****

It would be easy to think 'Surely we don't need another book on quantum physics.' There are loads of them. Anyone should be happy with The Quantum Age on applications and the basics, Cracking Quantum Physics for an illustrated introduction or In Search of Schrödinger's Cat for classic history of science coverage. Don't be fooled, though - because in Beyond Weird, Philip Ball has done something rare in my experience until Quantum Sense and Nonsense came along. It makes an attempt not to describe quantum physics, but to explain why it is the way it is.

Historically this has rarely happened. It's true that physicists have come up with various interpretations of quantum physics, but these are designed as technical mechanisms to bridge the gap between theory and the world as we see it, rather than explanations that would make sense to the ordinary reader.

Ball does not ignore the interpretations, though he clearly isn't happy with any of them. He seems to come clo…

Mercury - William Sheehan ****

Driving to work one morning several years ago, I spotted a tiny white dot close to the rising sun. ‘That’s Venus,’ I said to myself. Almost immediately I saw another, much brighter dot a few degrees away. ‘No, that’s Venus – the first one must be, um ... Mercury.’ Even with a lifelong interest in astronomy, I always manage to forget Mercury.

With eight planets in the Solar System, one of them has to be the least interesting – and Mercury got the short straw. That’s a relative statement, though, and a diligent author could still dig up enough fascinating facts about that tiny dot by the Sun to fill a short book. William Sheehan has done a brilliant job of doing just that.

One of the reasons Mercury is so easy to forget is that it’s almost impossible to get a good view of it from Earth. Even after the invention of the telescope, which turned planets like Mars and Jupiter into explorable worlds, Mercury remained a mystery – and the subject of some pretty wild speculations. In 1686, for exa…

Everything You Know About Planet Earth is Wrong - Matt Brown ****

This is the latest of a series of 'Everything You Know About... is Wrong' books from Matt Brown. Although I always feel slightly hard done by as a result of the assertion in the title, as there are certainly things here I know that aren't wrong (I mean, come on, the first corrected piece of 'knowledge' is that 'The Earth is only 6,000 years old' and I can't imagine many readers will 'know' that), it's a handy format to provide what are often surprisingly little snippets of information that are very handy for 'did you know' conversations down the pub (or showing up your parents if you're a younger reader).

Some of the incorrect statements that head each article are well-covered, if often still believed (for example, people thought that world was flat before Columbus), some are a little tricksy in the wording (such as seas have to wash up against land) and some are just pleasantly surprising (countering the idea that gold is a rar…