What you won't find here is any detail on the nature of quantum computers, how they work or on the very significant challenges faced in achieving anything that is to become mainstream. This is all treated at even higher level than a serious newspaper article would. What Michio Kaku is interested in is the potential applications, and the book takes us through a significant number of these.
You will read how quantum computers have the potential to transform our understanding of biology (physicists love telling biologists how they can do the job better), deal with green energy, transform medical cures, enhance AI, take a step towards immortality, predict and deal with with climate change and enable us to make far better cosmological models. It's heady stuff, and as always Kaku's viewpoint is both very positive and bolstered by an enthusiasm for science fiction (we even get a little SF story in 'a day in the Year 2050').
Of all those topics, for me by far the most interesting was the relatively short section on AI and quantum computing. AI is currently taking one of its occasional leaps forward with the likes of Chat-GPT, and Kaku looks at the ways in which AI and quantum computers could assist in dealing with each others weaknesses, enabling a transformational ability to make effective use of the vast amount of information now available to us online. Arguably, the book is worth reading for this alone.
However, there are some downsides to the approach taken. Kaku repeatedly overemphasises current capabilities. As the title suggests, we are told we have reached quantum supremacy, when quantum computers can far exceed the abilities of conventional computers. What is not made so clear is that this supremacy has only occurred with devices that are each focused on dealing with a very particular challenge, and that can't be applied to the real world in any way. This isn't just a failing of Kaku's, but the industry terminology. As he points out, in 2023 a senior figure at IBM said that within the next couple of years we will be able to 'reach a demonstration of quantum advantage - something that can have practical value.' Quantum advantage, which sounds trivial when compared with quantum supremacy, is actually far harder (and more important) to obtain - and even this statement isn't about general purpose quantum advantage. Kaku does not make clear at all how very specific quantum computing algorithms tend to be - these devices will never provide general purpose advantage.
The big problem here, and it's partly because we are given so little on how the computers work - and don't work - is the way that practically every application Kaku lists is 'may be' or 'might be able to'. We're a long way from knowing for sure how quantum computing can address more than a handful of requirements, such as calculating prime factors or speeding up search (the latter, of course, is a big prize for search engine companies). Even if quantum computers could address some of the issues Kaku lists - and we're never told how they would go about addressing them - it's not even always clear how the data could be collected to make their efforts usable.
Quantum computing is important. As Kaku points out, a huge amount of money has been invested in it, and it will become more practical for very specific applications, even if it's never commercially feasible outside of big data centres, so for most will be a cloud facility if used at all. And parts of some of Kaku's suggested applications will probably become reality. But the scattergun approach of describing possible applications without the practicalities doesn't really do justice to the issue.
Hardback:   |
Review by Brian Clegg - See all Brian's online articles or subscribe t o a weekly email free here
Comments
Post a Comment