This month's book reviews turn up a good primer for AI enthusiasts, an overview of politics in the power sector, and a little bit of zero-gravity how's your father.
Artificial Intelligence: the basics
by Kevin Warwick, £9.99, ISBN 978-0-415-56483-0
Whilst it's tempting to think that artificial intelligence (AI) is, if not new, then at least a digital-age phenomenon, the idea that non-humans might be able to think as we do has been around for centuries. Philosophers such as Descartes considered animals in terms of their machine performance, while the notion of artificial beings can be traced back to the stories of the Prague Golem, and even further to Greek mythology, such as Pygmalion's Galatea.
But the rise of what we now think of as AI has its classical roots much more recently and, as Kevin Warwick quite rightly observes in his book 'Artificial Intelligence: the Basics', it's practical to think of it as being notionally parallel with the development of the digital computer.
In fact, 'Classic AI' as he calls it – where from the 1950s scientists struggled to ascertain whether machines could think like humans – is as good a starting point as any. The problem is, in terms of technological thought, Classic AI is flawed in its thinking, simply because it is so human-centric, disallowing the possibility that machines might be capable of thought, but in a way entirely different from humans.
Warwick doesn't like this sort of anthropomorphism and is exasperated that there are still scientists and philosophers who think this way. Of course, it's fear, he says; fear that in some way machines might be better than us.
But, as he says, they are better in so many ways. When it comes to remembering things, calculating things and repeating data accurately, machines knock spots off even the finest of human minds. So why don't we further harness the intelligence of machines to take the drudgery out of human existence? As they get more intelligent, we'll have to learn to live alongside them, give them an equivalent to our human rights and respect them. For anyone thinking this is pie in the sky, read the book.
You won't find anything that hasn't been aired elsewhere, and if you know a lot about the subject you might want to concentrate on the research coming out of the University of Reading, where he is professor of cybernetics, working at the forefront of AI, control, robotics and biomedical engineering. But if AI is outside your field, or you know something of the subject and would like to know more, then 'Artificial Intelligence: the Basics' is a brilliant primer. Even though Warwick claims to have aimed not much higher than the intelligent 'A' level student or engineering undergraduate, in this I think he's doing himself and his book a disservice.
His key strength as an author is his ability to condense a huge field of intersecting philosophy, science and (let's face it) mumbo jumbo into a hugely enjoyable and comprehensive read. In fewer than 200 pages Warwick transmits the wider picture of current thinking in the world of AI in an agenda-neutral and yet highly enthusiastic way, while hinting at future directions. For some reason that I've never managed to work out, Warwick is a controversial figure in this field, but I suspect this book will make his detractors look at him in a new light.
Yale University Press
The Theory That Would Not Die
by Sharon Bertsch McGrayne, £18.99, ISBN 978-0300169690
Reverend Thomas Bayes came up with the idea of probability as 'degrees of belief' at a time when science and maths were the focus of intense debate in England about cause, effect, and the existence of God (considered to be the ultimate cause).
The philosopher David Hume's inflammatory essay of 1748 was one likely source of Bayes' inspiration, as was Newton's theory of gravitation. Hume said we cannot be sure about cause and effect, we can only rely on experience – and, crucially, the design of the world does not prove the existence of a creator. Newton had not explained the cause of gravity or looked at how true it might be but had instead reached his theory purely from observation.
Pierre-Simon Laplace derived what we today call Bayes's Rule independently – and then for the next two centuries, whether cracking Enigma codes, locating lost submarines or finding the causes of cancer, scientists and statisticians have fought over a deep philosophical divide about probability, which Sharon Bertsch McGrayne explores with great clarity and wit.
Using initial beliefs (a first guess about what might happen) plus recent data (concrete evidence) to generate an improved belief, Bayesians can quantify the chance of a single event like rain tomorrow; update hunches with new information; and include every scrap of data because each might change the answer by a small amount.
On the other side are those for whom probability must be the objective study of relative frequency. Dubbed the 'Frequentists', they think Bayes is subjectivity gone mad.
Bertsch McGrayne is such a good writer that she makes this obscure battle gripping for the general reader. By the time you reach the statistical gag on p129 ('How will you know one Bayesian from another? Ye shall know them by their posteriors') you will laugh at all levels of the joke, and not just the bottom reference.
Packing for Mars: The curious science of life in Space
by Mary Roach, £12.99(hb)/£8.99 (pb), ISBN 978-1-85168-780-0/823-4
According to journalist Mary Roach, 'Space is a world devoid of the things we need to live and thrive: air, gravity, hot showers, fresh veg, privacy, beer'. But 'what happens if you vomit in your helmet during a spacewalk?'. This astronaut exposé stuff has been done before, but not quite like this.
Roach has taken it upon herself to write about topics many shy away from, such as death and sex. With chapters such as 'The cadaver in the space capsule' and 'The three dolphin club: mating without gravity', one gets a sense of the author's favourite subjects.
Roach writes in the first person and makes herself an integral part of the story. She asks Jim Lovell what it smelt like in his space capsule; she describes photographs of herself in a wind tunnel as 'hideous' and she almost throws up in Tom Cruise's two-seat biplane. By the end of the book, however, the ego-trip gives way to 'getting it' about space, and holding a 9kg Mars meteorite 'does it' for her.
So, what about that dolphin club? It stems from writer G Harry Stine's assertion that clandestine, late-night experiments in a Nasa weightless simulation tank in the 1980s showed that sex is possible in microgravity. However, said Stine, 'it helped to have a third person to push at the right time in the right place [which] is the way dolphins do it'. A lot of rubbish is written about sex in space...and this is some of it, as Roach is well aware. But it all makes for a good story, and that's what this book is all about.
Power Politics: Political encounters in industry and engineering
by Francis Tombs, £24.50, ISBN 9781848855069
In an ideal world, one would like to think that the government had a duty to ensure the country enjoyed a secure and continuous supply of electricity; a duty second only to the security of the state. In this entertaining new book, Lord Tombs explains how, on the contrary, political decisions by successive governments since the Second World War 'have resulted in a situation where the reliability of electricity supply throughout the UK will be in serious jeopardy for many years to come'.
Lord Tombs is particularly well placed to make such a judgement. As chairman of the South of Scotland Electricity Board, and later chairman of the Electricity Council, his advice was sought (but seldom followed) by a succession of energy ministers who sought to reorganise the industry. Their method was apparently to rely on political dogma (sometimes nationalisation, sometimes privatisation) rather than technical knowledge and experience. Lord Tombs now sits as a cross-bench peer in the House of Lords.
The book starts with an account of how the reorganisation of the electrical supply industry was mishandled after the war. The construction of the National Grid in the 1920s and 30s, which allowed the nationwide transmission of electricity for the first time, had been a major achievement, replacing a system of expensive and inefficient local generation and distribution.
The author describes a 1950s Conservative government's decision to separate generation in England and Wales from distribution and sales as 'odd... entirely political, with little or no thought for the practical managerial consequences' .
Thus was created the Central Electricity Generating Board, which had a monopoly of generation until privatisation in 1988, and 12 local boards responsible for the separate job of distribution and sales. Lord Tombs identifies one consequence of this inefficient arrangement in that the CEGB promoted the construction of 'gold-plated' power stations which the contractors found were not competitive for overseas customers.
His recommendation was to implement a Bill that proposed the formation of an Electricity Corporation with a potential division into five autonomous and competing divisions each large enough to finance and build large power stations. Apparently the Conservative government of 1979 found it politically impossible to implement this proposal which had been favoured by the previous administration, and chose a different route for privatisation. One consequence has been that a number of UK electricity utilities have passed into foreign ownership.
The advent of cheap gas was, we find, a mixed blessing. Gas-fired power stations are quick and cheap to build but we have very little gas storage capacity to cover interruptions in supply and the reliance on one fuel, which the UK now has to import, has led to a decline in traditional industries building turbo-generators and combustion plant.
Lord Tombs is also critical of the 1997 Labour government's 'love affair' with wind power, which he describes as expensive to build and their potential value has been greatly exaggerated. He quotes the estimate of the subsidy required by wind power, £30bn by year 2020 ('a stealth tax in all but name') and says it is a sum more than sufficient to meet the cost of replacing the nuclear power stations.
He sees the 'visceral opposition of [Labour] to nuclear power and the accompanying obsession with wind power' as a threat to future supplies, and for the solution he says 'we sorely need a technically competent and independent body capable of long-term strategic planning for a national electricity supply'.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.