Article Source: The IET
You don’t need to be worried that a robot is coming for your job in order to be interested in the increasing role that intelligent machines are playing in our daily lives. We’ve rounded up a few books published this year that provide an authoritative view of what might be in store.
Warnings of how vulnerable different professions are to automation have become a popular theme during 2018. From the most low-skill tasks to sophisticated jobs like medical diagnostics, who wouldn’t want to know what the prospects are of a robot or clever piece of software replacing them in the months or years to come?
That popularity is reflected in the number of books that have come out in the past year, most of which – thankfully – take a more considered view than the click-hungry mainstream media. The Christmas break is a perfect time to sit back, get up to speed with what’s happening and perhaps plan around what the impact might be on your own career.
An excellent primer for the engineer interested in putting AI in context is ‘How Smart Machines Think’ by Sean Gerrish (The MIT Press, £22.00, ISBN 9780262038409). Rather than pontificating on the social implications of these innovations, Gerrish explains the key ideas that enable intelligent machines to perceive and interact with the world, describing, for example, the software architecture that allows self-driving cars to stay on the road and to navigate crowded urban environments; the million-dollar Netflix competition for a better recommendation engine; and how programmers trained computers to perform certain behaviours by offering them treats, as if they were training a dog. Computers haven’t yet mastered everything and Gerrish outlines the difficulties in creating intelligent agents that can successfully play video games, which have so far defied solution. If you’re the sort of person who, when a family member wonders out loud as they slump in front of the TV why movie recommendations don’t seem to suggest anything your household might actually be interested in watching, ‘How Smart Machines Think’ provides the ammunition you need to give them a comprehensive – if probably unwelcome – explanation.
Looking ahead decades rather than mere years, it’s likely that one day a driverless car will be able to navigate the roads better and more safely than you do and that algorithms will diagnose disease more reliably than human clinicians. Underpinning the ability of machines to mirror in years the evolution in intelligence that took humans millennia is deep learning, the technique fuelled by data in which networks learn in the same way that babies experience the world. Terrence J Sejnowski, who played an important role in the founding of deep learning in the 1980s, describes how information extracted from raw data can be used to create knowledge that in turn underlies understanding the understanding which leads to wisdom in ‘The Deep Learning Revolution’ (The MIT Press, £24, ISBN 9780262038034). In a story relating how deep learning went from being an arcane academic field to a disruptive technology in the information economy, Sejnowski explains it has brought us driverless cars, Google Translate, fluent conversations with the likes of Siri and Alexa, and enormous profits from automated trading on the New York Stock Exchange.
For those without the time to dive deep into AI, who want a physical book they can slip in their pocket to read on the train or waiting room, the ‘Very Short Introductions’ series from Oxford University Press is the perfect way to get ahead in a new subject quickly. One of the latest additions to the hundreds of titles already available is ‘Artificial Intelligence: A Very Short Introduction’ by Margaret A. Boden (Oxford University Press, £7.99, ISBN 9780199602919). Drawing on the applications of AI that we see all around us, Boden reviews the philosophical and technological challenges they raise, considering whether programs could ever be truly intelligent, creative or even conscious, and showing how their development has helped us to better appreciate the intricacies of human and animal minds.
Plenty of people are unconvinced by so called ‘techno-chauvinism’ – the assumption that technology is always the solution. They’ll argue that many of us are so eager to do everything digitally – hiring, driving, paying bills, even choosing partners – that we fail to recognise whether the technology actually works or not. ‘Artificial Unintelligence: How Computers Misunderstand the World’ (The MIT Press, £19.95, ISBN 9780262038003) is a techno-sceptic manifesto in which software developer and journalist Meredith Broussard argues that our collective enthusiasm for applying computer technology to every aspect of life has resulted in a tremendous number of poorly designed systems. Taking as her starting point the idea that there are fundamental limits to what we can and should do with technology, she offers a guide to understanding the inner workings and outer limits of technology based on a series of adventures in programming which includes an alarming ride in a driverless car. It’s only by understanding the limits of what we can do with technology, Broussard believes, that we can make better choices about what we should do with it to make the world a better place.
Believing that computers are smarter than us when they aren’t and trusting them to make vital decisions might be as dangerous as developing machines that genuinely can out-think humans; an idea echoed in Gary Smith’s ‘The AI Delusion’ (Oxford University Press, £20, ISBN 9780198824305). While he acknowledges that their ability to discover patterns means we can do things with computers that could never be done before, Smith makes the counterpoint that because they don’t think the same way that humans do, machines are useless at actually judging whether the unearthed patterns are sensible. Our concern should be less that super-intelligent machines will decide to protect themselves by enslaving or eliminating humans, he argues, and more that we’re being intimidated into thinking that computers are infallible and that black boxes should be trusted.
The question of whether or not we embrace AI isn’t an all or nothing one; it’s more about the extent to which it can complement the human attributes that it may never be able to emulate. Thomas W Malone, a professor of management at the MIT Sloan School of Management founding co-directors of an MIT initiative on ‘Inventing the Organisations of the 21st Century’, suggests in ‘Superminds: The Surprising Power of People and Computers Thinking Together’(OneWorld, £14.99, ISBN 9781786074713) that the best recipe will likely be teams in which humans and machines collaborate at just the right level. These aggregated ‘superminds’ are capable of acting more intelligently than any person, group or computer has ever done before and are in fact already making many of the world’s important decisions, whether through the hierarchy of a multinational company or the democracy of a nation state. Instead of fearing the rise of artificial intelligence, he concludes, we should be focusing on what we can achieve by working with computers.