Monthly Archives: March 2014

Welcoming the overlords

Posted on by

[amazon_link id=”023116856X” target=”_blank” ]Smart Machines: IBM’s Watson and the Era of Cognitive Computing[/amazon_link] is by a couple of IBM-ers, Director of Research John Kelly, and Steve Hamm. It shows. This short book is all about the promise of current massive increases in computers’ abilities rather than the disruption it will cause. It mentions Brynjolfsson and McAfee’s [amazon_link id=”0984725113″ target=”_blank” ]Race Against the Machine[/amazon_link], but it would be good to read this book alongside their new one, [amazon_link id=”0393239357″ target=”_blank” ]The Second Machine Age[/amazon_link]. With that caveat, it’s interesting to read this short book about what the insiders think is likely to happen.

[amazon_image id=”023116856X” link=”true” target=”_blank” size=”medium” ]Smart Machines: IBM’s Watson and the Era of Cognitive Computing (Columbia Business School Publishing)[/amazon_image]

The book starts with the famous victory by IBM’s Watson over two humans in the TV game show Jeopardy. One of the defeated men, Ken Jennings, famously said afterwards: “I for one welcome our new computer overlords.” The introductory section explains the key difference with the next generation of very powerful computers, the move away from the Von Neumann architecture – with its key bottleneck of moving data to and fro between CPU and memory – and instead having more parallel calculations with memory and processing more closely integrated. This will increase the machines’ capacity and enable them to go beyond specific programmes for specific tasks. They will be able to learn. Having just read Gerd Gigerenzer’s [amazon_link id=”0141015918″ target=”_blank” ]Gut Feelings[/amazon_link], though, I now feel a bit more optimistic about the complementarity between humans and robots – it’s hard to see how even smart machines can be taught or can acquire intuition given the way logic is integral to their structure and programming.

Later chapters of the book look at big data, at the new physics of computing – what future microchips could be like – and specifically at cities and the potential for cognitive computers to make them better places for their inhabitants. A key issue, given both the tendency for political power to shift from nation state to city level and the continuing move by humanity into cities. I suspect computer experts will be familiar with the material in this book, but I didn’t and it was an enjoyable travel read.

Enoughness

Posted on by

Yesterday I attended a very interesting symposium at Christ Church, Oxford on the 2012 book by Robert and Edward Skidelsky, [amazon_link id=”0241953898″ target=”_blank” ]How Much Is Enough? Money and the Good Life[/amazon_link]. I think my own [amazon_link id=”0691156298″ target=”_blank” ]The Economics of Enough [/amazon_link]was my passport to the event.

[amazon_image id=”0241953898″ link=”true” target=”_blank” size=”medium” ]How Much is Enough?: Money and the Good Life[/amazon_image]

The Skidelskys’ book is a good and provocative read. It starts with the question posed by the famous Keynes essay, [amazon_link id=”1441492267″ target=”_blank” ]Economic Possibilities for our Grandchildren[/amazon_link]: why are we not now enjoying much more leisure, given that Keynes was right in his prediction about the degree of technological progress. One aspect I very much agreed with is their point that it is important to keep the ideas of economic growth and well-being or social welfare distinct – this is one of the themes of my new book, [amazon_link id=”0691156794″ target=”_blank” ]GDP: A Brief But Affectionate History[/amazon_link] as well. GDP growth is a means to an end, reasonably highly correlated with well-being, but not an end in itself.

However, in other ways I strongly disagreed with their book. They argue for stopping growth and stopping technological progress, including, Lord Skidelsky explained yesterday, stopping or slowing down further automation of work by taxing the use of robots. To my mind, that completely misses the point about economic growth, which is not to have more and more of the same things, but to have more variety and new things. Innovation is the source of improvements in well-being, from trivial innovations to life-changing medicines or major technologies such as the internet and mobiles. GDP growth is the set of footprints left by innovation, although it fails to count the increase in welfare caused by new goods and services. At one point the Skidelskys write: “The material conditions for the good life already exist,” and say nothing has improved since 1974. This is absurd. Since 1974, UK life expectancy at birth has increased by 9 years to 81, the web has been invented, and we’ve moved from almost nobody having central heating to almost all of us having it – its lack is now taken as a poverty indicator.

The seminar included a long discussion about the a-morality or immorality of conventional economics. The economists taking part were probably atypical, as we all agreed that economics does need some post-crisis rethinking. The philosophers present focused on questions of liberalism and paternalism. Professor Cecile Fabre was particularly strong on questioning the Skidelskys’ failure to identify the ‘good life’ with some of the key values of liberalism.

I left with another book the two Skidelskys have edited, [amazon_link id=”0954643089″ target=”_blank” ]Are Markets Moral?[/amazon_link] Glancing at the contents, I think its answer is ‘no’. But even to pose the question is to make markets overly-abstract. Markets are institutions in which people have social relations. The markets for tea bags, accountancy services, and radio spectrum have entirely different structures and characteristics. ‘Moral’ isn’t a description that can apply to abstract nouns at all. One can sensibly ask if bankers are moral, at risk of generalising, but not markets.