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

The Dialogues - Clifford Johnson ***

The authors of science books are always trying to find new ways to get the message across to their audiences. In Dialogues, Clifford Johnson combines a very modern technique - the graphic novel or comic strip - with an approach that goes back to Ancient Greece - using a dialogue to add life to what might seem a dry message.

We have seen the comic strip approach trying to put across quite detailed science before in Mysteries of the Quantum Universe. As with that book, Dialogues manages to cover a fair amount of actual physics, but I still feel that the medium just wastes vast acres of page to say very little at all. This is brought home here because quite a lot of the sections of Dialogues start with several pages with no text on at all, just setting up the scenario.

As for using a discussion between two people to put a message across, Johnson makes the point that, for instance, Galileo's very readable masterpiece Two New Sciences is in the form of a dialogue (more accurately a discu…

Liam Drew - Four Way Interview

Liam Drew is a writer and former neurobiologist. he has a PhD in sensory biology from University College, London and spent 12 years researching schizophrenia, pain and the birth of new neurons in the adult mammalian brain. His writing has appeared in Nature, New Scientist, Slate and the Guardian. He lives in Kent with his wife and two daughters. His new book is I, Mammal.

Why science?

As hackneyed as it is to say, I think I owe my fascination with science to a great teacher – in my case, Ian West, my A-level biology teacher.  Before sixth form, I had a real passion for the elegance and logic of maths, from which a basic competence at science at school arose.  But I feel like I mainly enjoyed school science in the way a schoolkid enjoys being good at stuff, rather than it being a passion.

Ian was a revelation to me.  He was a stern and divisive character, but I loved the way he taught.  He began every lesson by providing us with a series of observations and fact, then, gradually, between …

A Galaxy of Her Own - Libby Jackson ****

This is an interesting book, even if it probably tries to be too many things to too many people. I wondered from the cover design whether it was a children's book, but the publisher's website (and the back of the book) resolutely refuse to categorise it as such. The back copy doesn't help by saying that it will 'inspire trailblazers and pioneers of all ages.' As I belong to the set 'all ages' I thought I'd give it a go.

Inside are featured the 'stories of fifty inspirational women who have been fundamental to the story of humans in space.' So, in some ways, A Galaxy of Her Own presents the other side of the coin to Angela Saini's excellent Inferior. But, inevitably, given the format, it can hardly provide the same level of discourse.

Despite that 'all ages' comment and the lack of children's book labelling we get a bit of a hint when we get to a bookplate page in the form of a Galaxy Pioneers security pass (with the rather worrying…