Skip to main content

The Master Algorithm - Pedro Domingos ***

I am really struggling to remember a book that has irritated me as much as this one, which is a shame because it's on a very interesting and significant subject. Pedro Domingos takes us into the world of computer programs that solve problems through learning, exploring everything from back propagating neural networks to Bayesian algorithms, looking for the direction in which we might spot the computing equivalent of the theory of everything, the master algorithm that can do pretty much anything that can be done with a computer (Turing proved a long time ago that there will always be some things that can't). As the subtitle puts it, this is the quest for the ultimate learning machine that will remake our world.

So far, so good. Not only an interesting subject but one I have a personal interest in as I had some involvement in artificial intelligence many moons ago. But just reading the prologue put my hackles up. It was one of those descriptions of how a technology influences every moment of your life, as the author takes us through a typical day. Except 90% of his examples have only ever been experienced by a Silicon Valley geek, and those that the rest of us have come across, like algorithms to make recommendations to you on shopping websites and video streaming sites, in my experience, are always so terrible that they are almost funny.

The pain carries on in part because of a kind of messianic fervour for the topic that means that the author seems convinced it is about to totally takeover the world - and like most fanatics, he presents this view while viciously attacking everyone who disagrees, from the likes of Marvin Minsky and Noam Chomsky to Black Swan author Nassim Nicholas Taleb. It's interesting that Domingos is totally dismissive of the early knowledge engineers who thought their methodology would take over the world, but can't see that his own pursuit of the 'master algorithm' (think of Lord of the Rings, but substitute 'algorithm' for 'ring') is equally likely to be a pursuit that is much easier to theorise about than to bring to success.

To make matters worse, Domingos repeatedly claims, for instance, that thanks to learning algorithms it's possible to predict the movement of the stock market, or to predict the kind of 'black swan' events that Taleb shows so convincingly are unpredictable. Yet I have never seen any evidence that this is true, it seems to go totally against what we know from chaos theory, and Domingos doesn't present any evidence, he just states it as fact. (Could you really have predicted the existence of black swans before they were discovered? How about blue ones?)

One other problem I have with the book is that the author isn't very good at explaining the complexities he is dealing with. I've seen many explanations of Bayesian statistics over the years, for instance, and this was one of the most impenetrable I've ever seen.

I can't tell you to avoid this book, because I've not come across another that introduces the whole range of machine learning options in the way that Domingos does. But any recommendation has to be made through gritted teeth because I did not like the way that information was put across.
Hardback:  

Kindle:  
Using these links earns us commission at no cost to you

These articles will always be free - but if you'd like to support my online work, consider buying a virtual coffee:
Review by Brian Clegg - See all Brian's online articles or subscribe to a weekly email free here
Labels:

Comments

Post a Comment

Popular posts from this blog

The Meteorite Hunters - Joshua Howgego *****

This is an extremely engaging read on a subject that everyone is aware of, but few of us know much detail about. Usually, if I'm honest, geology tends to be one of the least entertaining scientific subjects but here (I suppose, given that geo- refers to the Earth it ought to be astrology... but that might be a touch misleading). Here, though, there is plenty of opportunity to capture our interest. The first part of the book takes us both to see meteorites and to hear stories of meteorite hunters, whose exploits vary from erudite science trips to something more like an Indiana Jones outing. Joshua Howgego takes us back to the earliest observations and discoveries of meteorites and the initial doubt that they could have extraterrestrial sources, through to explorations of deserts and the Antarctic - both locations where it tends to be easier to find them. I, certainly, had no idea about the use of camera networks to track incoming meteors, which not only try to estimate where they wi...
Post a Comment
Read more

Phenomena - Camille Juzeau and the Shelf Studio ****

I am always a bit suspicious of books that are highly illustrated or claim to cover 'almost everything' - and in one sense this is clearly hyperbole. But I enjoyed Phenomena far more than I thought I would. The idea is to cover 125 topics with infographics. On the internet these tend to be long pages with lots of numbers and supposedly interesting factoids. Thankfully, here the term is used in a more eclectic fashion. Each topic gets a large (circa A4) page (a few get two) with a couple of paragraphs of text and a chunky graphic. Sometimes these do consist of many small parts - for example 'the limits of the human body' features nine graphs - three on sporting achievements, three on biometrics (e.g. height by date of birth) and three rather random items (GNP per person, agricultural yields of various crops and consumption of coal). Others have a single illustration, such as a map of the sewers of Paris. (Because, why wouldn't you want to see that?) Just those two s...
Post a Comment
Read more

Against the Odds - John Gribbin and Mary Gribbin ****

The number of women working in STEM subjects has expanded dramatically, but as John and Mary Gribbin make clear, in the history of science this is a very recent occurrence. Here, they bring us the stories of 12 women, from Eunice Newton Foote, born in 1819, to Vera Rubin, born in 1928 - effectively covering nearly 200 years in that Rubin died as recently as 2016. There are some names that will already be familiar from popular science histories (and deservedly so). You will find, for instance, Dorothy Hodgkin and Rosalind Franklin represented. But there are plenty like Foote that few will have come across, including Inge Lehmann, Chien-Sung Wu and Lucy Slater. While arguably Foote is there primarily to demonstrate the difficulties she faced (her discovery of an aspect of greenhouse gas behaviour was independently bettered within weeks), the rest have all made significant discoveries or developments against the odds and often missed out the recognition the deserved. The most prominent ob...
Post a Comment
Read more