The best science writers (as far as I'm concerned, at least) are the ones who make you feel like you're sitting down for a coffee with a smart friend, excited about a thing they know about, a thing they've found out, or a thing they've just put together.
I have never (yet) sat down for a coffee with Hannah Fry, but, having read Hello World, I can imagine I have. Finding the perfect balance between enthusiasm, expertise and clarity is a really tough trick, and - I note with a tinge of professional jealousy - she makes it look easy.
You can immediately hook a geek of a certain age1 with a well-placed reference to the ZX Spectrum; in this case, it's the first paragraph of the book, explaining the title. From there, the reader is taken on a tour of how computers - and specifically algorithms - bring their strengths and weaknesses to life in the 21st century.
It's divided into broad sections2 such as Justice, Health, Cars And Medicine, in each case discussing the triumphs and disasters of letting computers make decisions - without letting humans off the hook for being inconsistent and illogical.
And it's peppered with maths. I use the word 'peppered' advisedly: this is a book largely about the effects of maths, rather than the detail; the underlying techniques are the spice rather than the substance. All the same, it passes my "teach me something new" test (I'd often wondered what random forests were, and *Hello World* has saved me rooting out the enormous PDFs I downloaded with the best intentions some years ago); there are nice sections on the various types of error and Bayesian inference.
I loved it, and it made a long journey fly by; I only wonder if a better coach-routing algorithm could have got me there quicker (or at least found me a bus without a screaming infant).
* Disclosure: I was sent a copy of Hello World to review.