Artificial intelligence (also known as AI) refers to ‘intelligence’ exhibited by machines or software, as well as the industry that aims to develop these ‘intelligent’ machines. Underlying AI is the idea that machines can be ‘taught’ to adapt their behaviour to specific circumstances, thus achieving effective outcomes.
The idea of intelligent inanimate objects has long fascinated humans – a fact attested by the attempts of the ancient Egyptians to build automatons (nor were they the only culture of antiquity to try). In cultural terms, AI has cast a long shadow too – Mary Shelley’s Frankenstein, for example, may be seen as forewarning of the dangers of creating ‘artificial’ intelligent life, while movies like Terminator and I, Robot reflect our contemporary concerns.
Man versus machine
The modern concept of AI was arguably born at a conference at Dartmouth College, New Hampshire, in 1956, where the term was first used. One of the attendees that day, cognitive scientist Marvin Minsky, had high hopes. ‘Within a generation,’ he would predict, ‘the problem of creating “artificial intelligence” will substantially be solved.’ While that may have been over-optimistic, there have been high-profile successes in AI development over the years. In 1997, for instance, a computer (IBM’s Deep Blue) became the first machine to beat a chess champion, defeating Russian grandmaster Garry Kasparov. The computer had been trained to ‘think through’ the implications of each move better than even Kasparov could manage.
Yet there are deep moral and social implications to consider. What happens when machines are routinely more intelligent than us? What if they decide they should be in charge? Or what if their malice grows with their intelligence. In 2015, Stephen Hawking said:
The real risk with AI isn’t malice but competence. A super-intelligent AI will be extremely good at accomplishing its goals, and if those goals aren’t aligned with ours, we’re in trouble.
Another area of concern is the potential development of ‘autonomous weapons’ (sometimes known as ‘killer robots’) that could theoretically select and engage targets without any human involvement. War might conceivably be conducted by machines alone, erasing lines of human responsibility and removing opportunity for human mediation. Others fear that the robots will ultimately displace human workers on a mass scale – why employ a fallible and demanding human when a robot will do the job uncomplainingly for you? And what if a robot gets so smart that it decides simply to ignore, or even get rid of, its human overlords? Hawking’s suggested solution is to focus less on creating pure, undirected artificial intelligence in favour of what he calls ‘beneficial intelligence’ – machines that will only ever seek to serve our interests.