An exploration of the world experienced by a near-human artificial lifeform, trying to make her way and discover herself in an inimical world. It's not a new topic, of course. Since Asimov's robot stories in the 1950s there has been plenty of examination of this concept - and strictly it wasn't new then, as it's pretty much the theme of Pinocchio, dating back to 1883. This could hardly be more clear from the movie AI -Artificial Intelligence  (and to a lesser degree the Brian Aldiss short story it was based on, Supertoys Last All Summer Long ) where the puppet from Pinocchio is replaced by a synthetic human.  The fact this is a well-trodden path isn't a problem, though, because Paul Braddon manages to find new things to say and gives us an intriguing plot for Evie, the artificially intelligent creation in his novel. I was a bit worried for part of the first section of the book, which is quite slow, reflecting Evie's relatively limited life at that point, and set

Although Helgoland suffers from the usual issues Carlo Rovelli's books face - it is very short for the price and has a distinct tendency to purple prose - it is his best so far. In fact, the first hundred pages or so are excellent. Rovelli starts by giving us a brief background to quantum physics, concentrating most on Heisenberg, Schrödinger and to an extent Dirac's key period of contribution. This is clear and to the point. He then gives us a short summary of a couple of familiar quantum interpretations before introducing his own relational quantum physics interpretation. Although this idea dates back to the 1990s it has had very little coverage in popular science books, which is a shame.  Like all interpretations, Rovelli's requires us to accept some difficult postulates - in this case, that the reality of the state of a quantum system is relative rather than absolute - so it can be different for one observer than it is for another. Although at first this seems bizarre,

More like a similar number of bacteria One of my favourite things about being a science writer is getting emails from readers. Some say nice things about a book, others ask questions about science or writing which I do my best to respond to helpfully. However, every now and then there is an email that's pointing out an error in one of my books - and I have to admit that those are distinctly depressing. Let's be clear, I can say with reasonable confidence that every book I've ever written has a mistake or two in it. This isn't because I know what all the mistakes are - if I did, I would have corrected them before publication - but rather because pretty well every book I've ever read, fiction or non-fiction, has mistakes in it. It's just a fact of life. A book typically has between 60,000 and 120,000 words in it - the chances of an error slipping through is pretty high. Many of these are typos. Spelling mistakes, words missing, something that should be spotted in

There are two provisos here - first, this is primarily not a science book, but has enough science scattered through it to be worth covering. And second, I'm giving this book four stars because I enjoyed reading it despite its irritating flaws. It's a bit like a film I watched the other day in which an American actor did one of the worst Scottish accents I've ever heard. I still managed to enjoy the film, but I had to work at ignoring it. In Slow Rise , Robert Penn provides a memoir of his attempt to produce a year's worth of bread for his family using traditional wheat varieties he has grown and converted into flour himself. This may sound a bit like one of those amusing self-challenge books that Tony Hawks does so well, and there certainly is an element of humour in Penn's self-deprecating comments when things go wrong. But here this is also combined with some really interesting material on the biology and nature of wheat and the technical details of the bread-maki

Updated for paperback Things start well with this latest title from Brian Greene: after a bit of introductory woffle we get into an interesting introduction to entropy. As always with Greene's writing, this is readable, chatty and full of little side facts and stories. Unfortunately, for me, the book then suffers something of an increase in entropy itself as on the whole it then veers more into philosophy and the soft sciences than Greene's usual physics and cosmology. So, we get chapters on consciousness, language, belief and religion, instinct and creativity, duration and impermanence, the ends of time and, most cringe-making as a title, 'the nobility of being'. Unlike the dazzling scientific presentation I expect, this mostly comes across as fairly shallow amateur philosophising. This is a distinct case of attempting to discover the answer to 'life, the universe and everything', but missing the compactness of the answer 42. Of course it's perfectly possib

There is always a significant danger when a member of the literari takes on a science fiction theme - the result can easily seem derivative and dull when covering a topic that has already been better explored by others. (Of course the literary fiction audience are unlikely to realise this.) Sadly, there is an element of this danger manifest in Kazuo Ishiguro's new novel. The theme is a well-trodden one. A robot with strong artificial intelligence faces up to emotions and is used to explore the human condition. Here called 'artificial friends' we can see a progression to such companion robots from the current wave of cuddly robot pets, producing a device that has general artificial intelligence giving it the abilities and emotions of a human being. Klara is both a companion to and a replacement for a dying child. Of course there is nothing wrong with exploring a well-trodden path if you have something new to say, but most what occurs in Klara and the Sun is anything but ori