AP_I_Robots_1

This article was first published in the May 2016 China edition of Accounting and Business magazine.

Science fiction and films often create dystopian futures where androids, robots and other forms of artificial intelligence (AI) start out as helpers and then become a menace. Thanks to the likes of HAL 9000 in 2001: A Space Odyssey, VIKI in I, Robot, the ‘agents’ in The Matrix and Skynet in Terminator, we expect AIs to revolt against their human creators, take control of our society or pose an existential threat to the human race. 

Could this happen? In 2014, cosmologist and physicist Professor Stephen Hawking wrote in The Independent: ‘Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks.’ If an AI ever passes the Lovelace test (see ‘Testing for AI’, right) we may struggle to stop the worst from happening. Hawking says: ‘Humans, limited by slow biological evolution, couldn’t compete and would be superseded by AI.’ 

Can we do anything to prevent this? ‘I’m increasingly inclined to think there should be some regulatory oversight, maybe at the national and international level, just to make sure that we don’t do something very foolish,’ Elon Musk, tech visionary and CEO of Tesla Motors and SpaceX, told the AeroAstro Centennial Symposium at Massachusetts Institute of Technology. He said: ‘We should be very careful. With artificial intelligence, we are summoning the demon.’

One of Musk’s concerns is defence contractors and their creation of autonomous war robots that are designed to kill or wound humans. By comparison, other recent AI developments appear benign. Take Watson, the poster child for the incarnation of AI that the technology industry has labelled ‘cognitive computing systems’. IBM designed this to be ‘a natural extension of what humans can do at their best’. More of a helper than a menace.

Evolution or revolution?

When Watson defeated two titans of the US quiz show Jeopardy! in 2011, it captured the public imagination by appearing to understand the questions and coming up with the right answers faster than its human opponents – as if by magic. But appearances can be deceptive. To quote the writer and futurist Arthur C Clarke: ‘Any sufficiently advanced technology is indistinguishable from magic.’ Watson isn’t magic; it is a manifestation of the rapid evolution of computer systems.

The evolution of the human race took millions of years. It will not take this long for the non-biological intelligence of computer systems to evolve to match the range and subtlety of human intelligence. Processing power that required a room full of computers just decades ago now resides in your mobile phone. As processors have become smaller and more powerful, software developers have added more rules-driven engines to their products and these ‘expert’ systems have become faster and smarter.  

Watson demonstrated that a computer system can use ‘natural language processing’ software to recognise human speech, search huge amounts of complex information in seconds, generate and evaluate hypotheses, and adapt and learn through experience – and that was back in 2011. Since then, Watson has advanced and been joined by many other cognitive computer systems and so-called AIs. You may well have one on your mobile phone.

But are these systems ‘intelligent’? They can do things that require intelligence when carried out by humans, and they can fool us into thinking that they are also human (see ‘Testing for AI’). However, cognitive scientist professor Douglas Hofstadter says: ‘Watson is just a text search algorithm connected to a database, like Google search. It is finding text without having a clue what it means. In that sense, there is no intelligence there.’ Cognitive systems have their limitations.

Like children, Watson and other cognitive systems need educating. They can combine structured and unstructured data, then use context to determine the most appropriate response. But with a question such as ‘Which is the best tablet for me?’ the best response may depend on whether you are in a chemist or a computer store. Before cognitive systems can respond appropriately to such questions, they need training; they need to learn by adapting their future responses based on their past experiences. 

This appears significantly less threatening than gun-toting androids and self-replicating software entities with a consciousness and a sense of superiority. But even if today’s cognitive systems don’t pose an existential threat to humans (yet), they already look like bad news for some knowledge workers in some professions, such as accountancy. Perspectives differ on whether robotic process automation (RPA) is hype, helpful or horrifying. 

Beyond games

Meanwhile, Watson is still playing games. Edge Up Sports, for example, is developing a fantasy football app that uses Watson to analyse vast amounts of available football data. ‘By leveraging Watson technologies, we’re able to transform the way fantasy football is played, and provide a platform that is assisting team owners with analysis and insights that could increase their chances of winning their league,’ explains Illya Tabakh, CEO and head coach. 

This venture was made possible by IBM’s decision in 2014 to make Watson available in the cloud for general business use in a bid to make some serious money from it. But Watson is not just about playing games; it never was. Hundreds of ‘Watson Ecosystem Partners’ are using Watson’s smarts to create their own apps and recently released software tools that make non-IBM commercial development easier may encourage more developers to follow suit, inspiring all sorts of new products. 

IBM has said that it will target several key vertical industries, including retail, insurance and healthcare, and in April 2015 it launched its first official efforts to do so: IBM Watson Health and the Watson Health Cloud. Subsequent announcements have covered initiatives such as the development of a system at Boston Children’s Hospital to improve diagnosis of rare genetic conditions, and the first Watson deployment in India, where Manipal Hospitals will adopt it for oncology.

An eye on the future

IBM’s recent acquisition of Merge Healthcare Incorporated will give Watson the ability to see. ‘Watson’s powerful cognitive and analytical capabilities, coupled with those from Merge and other strategic acquisitions, position IBM to partner with organisations committed to changing the very nature of health and healthcare,’ says Dr John Kelly, a senior vice president at IBM Research. ‘Giving Watson “eyes” on medical images unlocks entirely new possibilities for the industry.’

Unsurprisingly, IBM and the partners in its Watson ecosystem are not the only ones excited by the possibilities of AI. While IBM has been exploring the possibilities of cognitive systems, so have numerous other technology companies. Google has become a frontrunner in the race from cognitive computing to AI. There are also plenty of other household names that are keen to exploit the commercial potential of systems that are smart enough to help us, without being smart enough to menace us. 

In October 2015, Facebook announced its new digital assistant: M. Facebook has taken a cognitive computing engine, trained it, and added it to the Facebook Messenger app. This digital assistant will search for information for you. If you give it permission, M will also oblige by assisting you with tasks in the physical world, such as making reservations for travel or restaurants, or making online purchases and arranging their delivery. So far, so helpful. 

Apple is also throwing money in the direction of AI. During 2015 it bought the British company Vocal IQ and its brand of cognitive computing, which can learn how you speak and use its ‘memory’ of previous conversations to improve each subsequent conversation. Sounds familiar? As well as buying AI technology that can improve machine processing of human speech, Apple has bought Perceptio, a specialist in image-recognition. The possibilities of cognitive systems and AI appear endless.

We will have to wait and see if cognitive systems lead to AI, and how well the human race deals with the consequences. Ginni Rometty, the chairman and CEO of IBM, is optimistic. ‘In the future, every decision that mankind makes is going to be informed by a cognitive system like Watson and our lives will be better for it,’ she says. However, back in 1943, namesake and IBM founder Thomas J Watson famously said: ‘I think there is a world market for maybe five computers.’

Lesley Meall, journalist