There have been many recent successes in the field of artificial intelligence. Some, like the 2011 success of IBM’s computer Watson beating champion human players in the popular TV show Jeopardy!, are ‘grandstanding’. Others are seemingly less obvious, but no less impressive for that. Face recognition, a task normally associated with humans and a few ‘higher-animals’, is now incorporated as routine by Facebook and Google’s Picasa. The ability to recognise speech is not far behind, as any user of the Siri feature on Apple iPhones can attest to. The point to note is that artificial intelligence technologies are becoming more important, more inconspicuous and an integral part of mainstream computing.
In The Intelligent Web, Gautam Shroff successfully argues that many of the recent successes have come through the deployment of many known but disparate techniques working together. He further argues that these advances have come about by this deployment being on large volumes of ‘big data’; all of which has been made possible , and indeed driven by the internet and the world wide web. In other words, Shroff argues, rather than ‘traditional’ artificial intelligence, these successes are best described as ‘web intelligence’. Semantic arguments aside, Shroff presents his case logically and convincingly.
What Shroff argues is that the cumulative use of artificial intelligence techniques at web scale can result in behaviour that exhibits the very basic feature of human intelligence, to; ‘put two and two together’ or ‘connect the dots’. In the book he dissects the ability to connect the dots. He argue, quite convincingly, that this ability is composed of looking and experiencing the world around us, next is to listen to what is important and discard the irrelevant, then to connect different facts and derive new conclusions. Then is the ability to to make predictions about the future and finally go put these to good use to correct and control our actions.
The bulk of the book is chapters devoted to each of these steps. Shroff accomplishes this with great diligence and intelligence. each chapter is replete with explanation and useful, if not energising, examples. the epilogue acknowledges that tricky question regarding any discourse on artificial intelligence, ‘where do the goals come from?” “Obviously the answer” says Shroff, “is from us. The web intelligence systems themselves do not generate their overarching goals; these are built in, by their creators, i.e., us.”
The missing ingredient between these ‘artificial intelligence’ systems and us, as ‘natural intelligence’ systems is purpose. Shroff in his epilogue poses the thought that for the web intelligence systems of today to cross the chasm, integrate these six elements and become a mind is an achievable scientific task worthy of the effort. As he presents the problem it seems far more close than it did in Turing’s time, when even the possible path was mere whimsy.
At the same time this book is about the ‘weak’ artificial intelligence. Shroff avoids the deep philosophical waters of strong artificial intelligence; whether or not machines can be conscious or ‘truly intelligent’. The tenuous links are there for the reader to make if they choose. As presented by Shroff these links are more tenuous than the very worthwhile and entertaining Wetware: a computer in every living cell by Dennis Bray. Where Bray provocatively engages in the debate about the computational capabilities of protein networks (not too dissimilar from an electronic web network).
Despite the success of his arguments there is a certain cumbersome feel to Shroff’s writing. In the prologue of his book Shroff says how he “grew up reading and being deeply influenced by the popular science books of George Gamow” and “I hope to entertain and elucidate, in the spirit of Gamow.” This rather lofty goal Shroff does fall short on. This book requires attention and diligence to read. It does not effortlessly tackle a complex subject with the insightful ramble of Bray in Wetware, for example. It is worth the effort, but I doubt it will be remembered with the same ‘affection’ as, for example, Gamow’s Mr Tomkins in Wonderland.
This review was first published here: [...]