Skip to main content

How Smart Machines Think - Sean Gerrish ****

While it will become apparent I think this book should have been titled 'How Dumb Machines Think', it was a remarkably enjoyable insight into how the well publicised AI successes - self-driving cars, image and face recognition, IBM's Jeopardy! playing Watson, along with  game playing AIs in chess, Go and Atari and StarCraft, perform their dark arts.

There's no actual programming presented here, so no need for non-programmers to panic, though there is some quite detailed discussion of how the software architectures are structured and how the different components - for example neural networks - do their job, but it isn't anything too scary if you take it slowly.

One thing that comes across very strongly, despite the AI types' insistence that their programs are of general use, is how very specifically tailored programs like the AlphaGo software that beat champions at the game Go, and the Watson computer that won at the US TV quiz show Jeopardy! were - incredibly finely designed to meet use and that use only.

The reason I make the remark about dumb machines is that what doesn't come across sufficiently in Sean Gerrish's book is that, because these programs are not in any sense intelligent, when they get things wrong, they often get things dramatically wrong. So some of Watson's answers on Jeopardy! did not make any sense at all. Similarly, image recognition software can be fooled by apparently abstract patterns that happen to have the right components to appear to be a distinguishable object. And when you bear in mind we're suggesting putting this kind of software in charge of cars that 'getting it dramatically wrong' bit is more than a little unnerving.

There was, though, a great section on the development of self-driving cars, from the original feeble attempts, where all the competitors in a race failed before completing 10 percent of the course, through to more recent and more successful versions that can handle basic traffic scenarios - though it would have been nice if Gerrish had gone beyond the old prize challenges to describe what the latest Google and Uber vehicles do. (It may be that their approaches are too proprietary.)

However, what was missing was any serious assessment of the big problems still faced. There have been two excellent books recently on the huge holes in AI that practitioners rarely admit to - The AI Delusion and Common Sense, The Turing Test and the Quest for Real AI - Gerrish would have produced an even better book if he could have addressed the concerns that these books raise.

Even so, in How Smart Machines Think we have a hugely informative and very readable book for anyone with an interest in finding out just what the much-trumpeted AI systems really do, and what lies beneath the hype.


Using these links earns us commission at no cost to you

Review by Brian Clegg


Popular posts from this blog

Tim Woollings - Four Way Interview

Tim Woollings is an Associate Professor in Physical Climate Science at the University of Oxford, leading a team of researchers in the Atmospheric Dynamics group. He obtained his PhD in Meteorology in 2005 and since then has worked on a variety of topics spanning weather prediction, atmospheric dynamics and circulation, and the effects of climate change. He has studied how the jet stream varies over weeks, years, and decades, and how we can better predict these changes. He was a contributing author on three chapters of the IPCC Fifth Assessment Report. Tim worked at the University of Reading as a postdoc, research fellow and then lecturer before moving to the University of Oxford in 2013. He is now the Oxford Joint Chair of the Met Office Academic Partnership. His new book is Jet Stream.

Why climate?

It has never been more important to learn about how our climate system works, and how we are affecting it. You certainly get a lot of satisfaction when your work touches on hugely important …

The Apollo Chronicles - Brandon Brown *****

There were two reasons I wasn't expecting much from this book. Firstly, there have been so many titles on the Apollo programme and the space race. And secondly, a book that focusses on the engineering involved would surely be far too much at the nut and bolt level (literally), missing out on the overarching drama that makes the story live. Also there were so many people involved - 400,000 is mentioned - that we couldn't have much human interest because we would be bombarded with lists of names.

Instead, I was charmed by Brandon Brown's account. His father was one of the engineers, but he isn't given undue prominence - Brown picks out a handful of characters and follows them through, bringing in others as necessary, but never overwhelming us with names. And while it's true that there is a lot of nitty gritty engineering detail, it rarely becomes dull. Somehow, Brown pulls off the feat of making the day-to-day, hectic engineering work engaging.

I think in part this …

Saturn – William Sheehan ****

This book marks something of a milestone in my reviewing career: it’s the first time I’ve seen an excerpt from one of my reviews printed on the back cover. It comes from my review of Sheehan’s previous book, on Mercury, which I said ‘easily convinced me the Solar System’s 'least interesting' planet is still a pretty fascinating place.’ That wasn’t an easy task for the author, given Mercury’s unspectacular appearance and reputation – but Saturn is a different matter. With its iconic rings, easily visible through a small telescope, it’s the favourite planet of many amateur astronomers. For space scientists, too, it’s a prime target – given that two of its moons, Titan and Enceladus, look like the kind of places we might find alien life. So Sheehan’s challenge this time wasn’t to find enough material to fill 200 pages, but to distil a potentially huge subject down to that size.

He meets this challenge just as successfully as the previous one – but not quite in the way I was expect…