Algorithms versus spontaneity – striking the happy medium
There’s been a number of recent pieces about the rise of self-learning technology that uses artificial intelligence (AI) to carry out tasks that would previously have been too complex for a machine. From stock trading to automated translations and even playing Frogger, computers will increasingly take on roles that used to rely on people’s skills.
Netflix used an algorithm to analyse the most watched content on its service, and found that it included three key ingredients – Kevin Spacey, director David Fincher and BBC political dramas. So when it commissioned original content, it began with House of Cards, a remake of a BBC drama, starring Spacey and directed by (you’ve guessed it) Fincher.
This rise of artificial intelligence is worrying a lot of people – and not just Luddites. The likes of Stephen Hawking, Bill Gates and Elon Musk have all described it as a threat to the existence of humanity. They worry that we’ll see the development of autonomous machines with brains many thousands of times larger than our own, and whose interests (and logic) may not square with our own. Essentially the concern is that we’re building a future generation of Terminators without realising it.
They are right to be wary, but a couple of recent stories made me think that human beings actually have several big advantages – we’re not logical, we don’t follow the facts and we don’t give up. Psychologist Daniel Kahneman won a Nobel Prize for uncovering the fact that the human mind is made up of two systems, one intuitive and one rational. The emotional, intuitive brain is the default for decision making – without people realising it. So in many ways AI-powered computers do the things we don’t want to do, leaving us free to be more creative (or lazy, dependent on your point of view).
Going back to the advantages that humans have over systems, the first example I’d pick is the UK general election. All the polls predicted a close contest, and an inevitable hung parliament – but voters didn’t behave logically or according to the research and the Tories trounced the opposition. While you might disagree with the result, it shows that you can’t predict the future with the clarity that some expect.
Humans also have an in-built ability to try and game a system and find ways round it, often with unintended consequences. This has been dubbed the Cobra effect after events in colonial India. Alarmed by the number of cobras on the loose, the authorities in Delhi offered a bounty for every dead cobra handed in. People began to play the system, breeding snakes specifically to kill and claim their reward. When the authorities cottoned on and abandoned the programme, the breeders released the now worthless snakes, dramatically increasing the wild cobra population. You can see the same attempt to rig the system in the case of Navinder Singh Sarao, the day trader who is accused of causing the 2010 ‘flash crash’ by spoofing – sending sell orders that he intended to cancel but that tricked trading computers into thinking the market was moving downwards. Despite their intelligence, trading systems cannot spot this sort of behaviour – until it is obviously too late.
The final example is when humans simply ignore the odds and upset the form book. Take Leicester City. Rock bottom of the English Premiership, the Foxes looked odds-on to be relegated. Yet the players believed otherwise, kept confident and continued to plug away. The tide now looks as if it has turned, and the team is just a couple of points away from safety. A robot would have long since given up……..
So artificial intelligence isn’t everything. Giving computers the ability to learn and process huge amounts of data in fractions of a second does threaten the jobs of workers in the knowledge economy. However it also frees up humans to do what they do best – be bloody minded and subversive, think their way around problems, and use their intuition rather than the rational side of their brain. And of course, computers still do have an off switch………….