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

Comments

Popular posts from this blog

The Decline and Fall of the Human Empire - Henry Gee ****

In his last book, Henry Gee impressed with his A (Very) Short History of Life on Earth - this time he zooms in on one very specific aspect of life on Earth - humans - and gives us not just a history, but a prediction of the future - our extinction. The book starts with an entertaining prologue, to an extent bemoaning our obsession with dinosaurs, a story that leads, inexorably towards extinction. This is a fate, Gee points out, that will occur for every species, including our own. We then cover three potential stages of the rise and fall of humanity (the book's title is purposely modelled on Gibbon) - Rise, Fall and Escape. Gee's speciality is palaeontology and in the first section he takes us back to explore as much as we can know from the extremely patchy fossil record of the origins of the human family, the genus Homo and the eventual dominance of Homo sapiens , pushing out any remaining members of other closely related species. As we move onto the Fall section, Gee gives ...

The New Lunar Society - David Mindell *****

David Mindell's take on learning lessons for the present from the eighteenth century Lunar Society could easily have been a dull academic tome, but instead it was a delight to read. Mindell splits the book into a series of short essay-like chapters which includes details of the characters involved in and impact of the Lunar Society, which effectively kick-started the Industrial Revolution, interwoven with an analysis of the decline of industry in modern twentieth and twenty-first century America, plus the potential for taking a Lunar Society approach to revitalise industry for the future. We see how a group of men (they were all men back then) based in the English Midlands (though with a strong Scottish contingent) brought together science, engineering and artisan skills in a way that made the Industrial Revolution and its (eventual) impact on improving the lot of the masses possible. Interlaced with this, Mindell shows us how 'industrial' has become something of a dirty wo...

Pagans (SF) - James Alistair Henry *****

There's a fascinating sub-genre of science fiction known as alternate history. The idea is that at some point in the past, history diverged from reality, resulting in a different present. Perhaps the most acclaimed of these books is Kingsley Amis's The Alteration , set in a modern England where there had not been a reformation - but James Alistair Henry arguably does even better by giving us a present where Britain is a third world country, still divided between Celts in the west and Saxons in the East. Neither the Normans nor Christianity have any significant impact. In itself this is a clever idea, but what makes it absolutely excellent is mixing in a police procedural murder mystery, where the investigation is being undertaken by a Celtic DI, Drustan, who has to work in London alongside Aedith, a Saxon reeve of equivalent rank, who also happens to be daughter of the Earl of Mercia. While you could argue about a few historical aspects, it's effectively done and has a plot...