Skip to main content

Algorithms to Live By - Brian Christian and Tom Griffiths *****

I was captivated by much of this book. It's the perfect antidote to the argument you often hear from young maths students - 'What's the point? I'll never use this in real life!' This often comes up with algebra (which often is useful), but reflects the way that we rarely cover the most applicable bits of maths to everyday life at high school. Although this book is subtitled 'the computer science of human decisions', it's really about the maths of human decision making (which is often supported by computers) - I suspect the 'computer science' label is to make it more sexy than boring old mathematics.

If there is any danger that the 'M' word would turn you off, the book tends to skip over the mathematical workings, concentrating on the outcomes and how they're relevant to the kind of decisions we make in everyday life - and it's that application side that makes it particularly interesting (helped by a good, readable style from the co-authors). So, for instance, one of the earliest areas covered is the kind of decision where you are selecting between a number of options that arrive sequentially and where you have to make a decision on which is best for you part way through the sequence, even though there may be better options in the future. The classic examples for this are some kinds of job interviews, house buying and finding a partner for life.

It might seem there can be no sensible advice, but mathematically it's very clear. You wait until you've got through 37% of the choices, then pick the next one that's better than any you've seen before. It's not that this will necessarily deliver your best of all possible worlds. More often than not it won't. But it will give you a better result than any other mechanism for deciding when to go for a particular option. Of course it's not always easy to apply. For example, unless it's something like an interview with closed applications, how do you know when you are 37% of the way through the available options? Luckily, the authors point out that there are approximations to get around this, which include that the approach can also apply to the amount of time available for the process.

And that's just the start. Along the way you will discover the best way to sort the books on your shelves into alphabetic order (something I confess I did last year, using a sub-optimal mechanism), how to balance exploration (for example, trying out new restaurants) with exploitation (for example, returning to tried and tested restaurants), how the concept of caching can revolutionise your filing system (and make that pile of papers on your desk that everyone mocks the sensible approach), why Bayes theorem is so important and much more. I absolutely revelled in this book.

The content only fades a bit when the applications aren't about real world decisions. So, for instance, there's some material about how the internet works that is very interesting if you like that kind of thing (I do), but hasn't got the same feeling of personal utility to it, so lacks some of the bite of the other chapters. This is even more obvious in the section on randomness. I would also have liked to see more acknowledgement that most of the content was really from the area of study called operational (operations in the US) research, a discipline that happens to make use of computers, rather than true computer science - but that's a specialist moan.

Realistically speaking, I don't think much of the content of this book will truly change how any of us do things. Interestingly, the authors reveal than an expert in the field pretty much consciously ignored the mathematical approach in a particular case, opting for more of a 'feels right' choice. But that doesn't stop the whole business, whether it's the relative simplicity of the 37% rule or the mind twisting possibilities of game theory, from being both potentially practical and highly enjoyable as presented here. Recommended.

Hardback 

Kindle 
Review by Brian Clegg

Comments

Popular posts from this blog

The Art of Statistics - David Spiegelhalter *****

Statistics have a huge impact on us - we are bombarded with them in the news, they are essential to medical trials, fundamental science, some court cases and far more. Yet statistics is also a subject than many struggle to deal with (especially when the coupled subject of probability rears its head). Most of us just aren't equipped to understand what we're being told, or to question it when the statistics are dodgy. What David Spiegelhalter does here is provide a very thorough introductory grounding in statistics without making use of mathematical formulae*. And it's remarkable.

What will probably surprise some who have some training in statistics, particularly if (like mine) it's on the old side, is that probability doesn't come into the book until page 205. Spiegelhalter argues that as probability is the hardest aspect for us to get an intuitive feel for, this makes a lot of sense - and I think he's right. That doesn't mean that he doesn't cover all …

The Best of R. A. Lafferty (SF) – R. A. Lafferty ****

Throughout my high school years (1973–76) I carefully kept a list of all the science fiction I read. I’ve just dug it out, and it contains no fewer than 1,291 entries – almost all short stories I found in various SF magazines and multi-author anthologies. Right on the first page, the sixth item is ‘Thus We Frustrate Charlemagne’ by R. A. Lafferty, and his name appears another 32 times before the end of the list. This isn’t a peculiarity of my own tastes. Short stories were much more popular in those days than they are today, and any serious SF fan would have encountered Lafferty – a prolific writer of short fiction – in the same places I did.

But times change, and this Gollancz Masterworks volume has a quote from the Guardian on the back describing Lafferty as ‘the most important science fiction writer you’ve never heard of’. Hopefully this newly assembled collection will go some way to remedying that situation. It contains 22 short stories, mostly dating from the 1960s and 70s, each w…

David Beerling - Four Way Interview

David Beerling is the Sorby Professor of Natural Sciences, and Director of the Leverhulme Centre for Climate Change Mitigation at the University of Sheffield. His book The Emerald Planet (OUP, 2007) formed the basis of a major 3-part BBC TV series ‘How to Grow a Planet’. His latest title is Making Eden.

Why science?

I come from a non-academic background. None of my family, past or present, went to university, which may explain the following. In the final year of my degree in biological sciences at the University of Wales, Cardiff (around 1986), we all participated in a field course in mid-Wales, and I experienced an epiphany. I was undertaking a small research project on the population dynamics of bullheads (Cotus gobio), a common small freshwater fish, with a charismatic distinguished professor, and Fellow of the Royal Society in London. Under his guidance, I discovered the process of learning how nature works through the application of the scientific method. It was the most exciting t…