Skip to main content

The Golden Ticket: P, NP, and the Search for the Impossible – Lance Fortnow ***

There is good and bad news early on in this book about the P versus NP problem that haunts computing. The good news is that on the description I expected this to be a dull, heavy going book, and it’s not at all. Lance Fortnow makes what could be a fairly impenetrable and technical maths/computing issue light and accessible.
The bad news is that frustratingly he doesn’t actually tell you what P and NP mean for a long time, just gives rather sideways definitions of the problem along the lines of ‘P refers to the problems we can solve quickly using computers. NP refers to the problems to which we would like to find the best solution’, and also that he makes a couple of major errors early on, which make it difficult to be one hundred percent confident about the rest of the book.
The errors come in a section where he imagines a future where P=NP has been proved. This would mean you could write an algorithm to very efficiently match things and select from data. Fortnow suggests that our lives would be transformed. This is slightly cringe-making as fictional future histories often are, but the real problem is that he tells us that the algorithm would make it possible to do two things that I think just aren’t true.
First he says that from DNA you would be able to identify what a person looks like and their personality. Unfortunately, these are both strongly influenced by epigenetic/environmental issues. Anyone who knows adult identical twins (with the same basic DNA) will know that they can look quite different and certainly have very different personalities. And they will usually have been brought up in the same environment. Fortnow is forgetting one of the oldest essentials of computing – it doesn’t matter how good your algorithm is, GIGO – garbage in; garbage out.
The other, arguably worse error is that he says that it will be possible to have accurate weather forecasts going forward X days. This is so horribly wrong. He should have read my book Dice World. The reason you can’t predict the weather at all beyond about 10 days is nothing to do with the quality of the model/algorithm, it is because the system is chaotic. Firstly we just don’t know, and never can know, the initial conditions to enough decimal places not to deviate from the real world. When Lorenz first discovered chaos it was because he entered the starting values in his model to 4 decimal places rather than the 6 to which the model actually worked. It soon deviated from the previous run. We can’t measure things accurately enough. The other problem is that the weather system is so complex – hence the slightly misleading title of Lorenz’s famous paper Does the flap of a butterfly’s wings in Brazil set off a tornado in Texas? – that we can’t possible take into account enough inputs to ever have so good a model as to go forwards that far. Sorry, Lance, it ain’t going to happen.
For the rest, the first half or so of the book goes along pretty well, gradually opening up the nature of P and NP, the problems that are of interest and the ‘hardest’ NP complete problems. I found the main example, used throughout, a hypothetical world called Frenemy where everyone is either a friend or enemy of everyone else confusing and not particularly useful, but Fortnow gets plenty of good stuff in. After that it’s as if he rather runs out of material and it gets a bit repetitious or has rather tangential chapters.
Overall, despite the flaws, a much better and more readable book than I thought it was going to be – but probably best for maths/computing buffs rather than the general popular science audience.

Hardback 

Kindle 
Review by Brian Clegg

Comments

Popular posts from this blog

Rockets and Rayguns - Andrew May ****

The Cold War period saw dramatic developments in science and technology, coinciding with the flourishing of the science fiction genre. In Rockets and Rayguns, Andrew May draws on the parallels between reality and fiction, each influencing the other.

Inevitably a major Cold War theme was the threat of nuclear war, and May opens with the bomb. It's fascinating that fiction got there first - nuclear weapons were featured in science fiction when many physicists were still doubting the practicality of using nuclear energy. Of course, it's a lot easier to simply take a concept and dream up a weapon than it is to make it for real - for example, H. G. Wells' prophetic nuclear bombs from his 1914 The World Set Free were nothing like the real thing. And some science fiction devices concepts - notably ray guns and force fields - came to very little in reality. However this doesn't prevent the parallels being of interest.

May gives us a mix of the science - describing how nuclear we…

Galileo Galilei, the Tuscan Artist – Pietro Greco ****

Near the beginning of John Milton’s epic poem Paradise Lost, he refers to a ‘Tuscan artist’ viewing the Moon through an optic glass. He’s talking about Galileo – one of history’s greatest scientists, but not the most obvious person to slap an ‘artist’ label on. Yet Galileo lived at a time – the Renaissance – when it was fashionable to dabble impartially in both the arts and sciences. Look up ‘Renaissance man’ on Wikipedia and you’ll see Galileo’s picture right there underneath Leonardo da Vinci’s. It’s a less well-known side to his life, but it crops up again and again – interspersed among his many scientific achievements – in this excellent new biography by Pietro Greco.

If you’re looking for interesting trivia, you’ll find plenty in this book. Galileo’s father was a musician with scientific leanings, who carried out some of the first experiments on musical acoustics – which Galileo may have assisted with. As a young professor of mathematics, Galileo delivered a couple of lectures on …

The Ultimate Interplanetary Travel Guide - Jim Bell **

Not too long ago, NASA brought out a series of spoof ‘space tourism’ posters for various destinations in the Solar System. For people like me, who have watched NASA depressingly fail to send humans to the other planets for decade after decade, it was just another painful twist of the knife. On the other hand, the posters probably have more appeal for younger and less cynical minds, by presenting familiar astronomical objects in a new and engaging way. There may even be scope for a whole book along these lines – and that’s what Jim Bell has attempted here.

The main thing I learned from it is that my brain is programmed to read either fiction or non-fiction, and can’t handle a 50:50 mix of the two – which is what this book is. It drove me mad –  not least because I could see that the same material, presented as straight non-fiction, could have made a really excellent book. Using the NASA posters as a starting point, he could have enumerated the potential ‘tourist sights’ at each location…