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:  
Using these links earns us commission at no cost to you
Review by Brian Clegg
Labels:

Comments

Post a Comment

Popular posts from this blog

The Antigravity Enigma - Andrew May ****

Antigravity - the ability to overcome the pull of gravity - has been a fantasy for thousands of years and subject to more scientific (if impractical) fictional representation since H. G. Wells came up with cavorite in The First Men in the Moon . But is it plausible scientifically?  Andrew May does a good job of pulling together three ways of looking at our love affair with antigravity (and the related concept of cancelling inertia) - in science fiction, in physics and in pseudoscience and crankery. As May points out, science fiction is an important starting point as the concept was deployed there well before we had a good enough understanding of gravity to make any sensible scientific stabs at the idea (even though, for instance, Michael Faraday did unsuccessfully experiment with a possible interaction between gravity and electromagnetism). We then get onto the science itself, noting the potential impact on any ideas of antigravity that come from the move from a Newtonian view of a...
Post a Comment
Read more

The World as We Know It - Peter Dear ***

History professor Peter Dear gives us a detailed and reasoned coverage of the development of science as a concept from its origins as natural philosophy, covering the years from the eighteenth to the twentieth century. inclusive If that sounds a little dry, frankly, it is. But if you don't mind a very academic approach, it is certainly interesting. Obviously a major theme running through is the move from largely gentleman natural philosophers (with both implications of that word 'gentleman') to professional academic scientists. What started with clubs for relatively well off men with an interest, when universities did not stray far beyond what was included in mathematics (astronomy, for instance), would become a very different beast. The main scientific subjects that Dear covers are physics and biology - we get, for instance, a lot on the gradual move away from a purely mechanical views of physics - the reason Newton's 'action at a distance' gravity caused such ...
Post a Comment
Read more

It's On You - Nick Chater and George Loewenstein *****

Going on the cover you might think this was a political polemic - and admittedly there's an element of that - but the reason it's so good is quite different. It shows how behavioural economics and social psychology have led us astray by putting the focus way too much on individuals. A particular target is the concept of nudges which (as described in Brainjacking ) have been hugely over-rated. But overall the key problem ties to another psychological concept: framing. Huge kudos to both Nick Chater and George Loewenstein - a behavioural scientist and an economics and psychology professor - for having the guts to take on the flaws in their own earlier work and that of colleagues, because they make clear just how limited and potentially dangerous is the belief that individuals changing their behaviour can solve large-scale problems. The main thesis of the book is that there are two ways to approach the major problems we face - an 'i-frame' where we focus on the individual ...
Post a Comment
Read more