Monday, 8 December 2014

Run away! The robots are coming! Aargh!

No lesser personage than Steven Hawking has announced the end of humanity as we know it if it pursues its desire for ever more intelligent machines. Once they reach the level of human insight and analysis, he claims, they will start to communicate with one another without our even knowing it and eventually, as did the computer Hal in the film 2001: A Space Odyssey, conclude that humans are no longer necessary for them to carry out their mission, whatever that is. This view exactly mirrors the plotline in The Terminator movies, where computers from the future have actually declared war on the human race.


Now I would normally be slightly reluctant to take issue with one of the most intelligent people alive on this planet. Steven Hawking is said to have an IQ of 215; hence it would probably be unwise to argue with him about anything. However, reckless as it may be, I do choose to take issue with him on this point.


Alan Turing is hot right now, with Benedict Cumberbatch (who is even hotter) playing him in the new film The Imitation Game. I haven't seen it yet, but I wonder if, along with the struggle to crack the Enigma code of which Turing was a critical part, there is any discussion of one of his most famous concepts: "The Turing Test". Trying to answer the question "Can computers think, or could they ever think?" he proposed this acid test: if by communicating with a computer, by whatever means, their response is indistinguishable from how an intelligent human being behind a screen might respond, then to all intents and purposes that computer is thinking.


No computer has yet passed the Turing test, but around the world the race is on to design software that could do it, and when it is finally achieved, it will be one of the greatest scientific breakthroughs of all time. That great futurologist Arthur C Clarke imagined his computer Hal could pass the Turing test with ease, and that by the year 2001. Thirteen years past that date it still hasn't happened, but I am told that the computer world is finally closing on this holy grail. In some ways there are already signs we are approaching it, like when you mis-type something into google and it helpfully asks if you meant something else, which is usually exactly what you did mean.


I might live long enough to see this historic milestone passed, and it will certainly change the world irrevocably. Apart from anything else, it could end loneliness once and for all, as we could purchase a sentient companion we could talk to, even have a meaningful and caring relationship with. Indeed, there is perhaps no limit to what could be achieved. Obviously safeguards would need to be built in (see Asimov's three rules of robotics)- but wouldn't it be nice if we could build similar safeguards into our relationships with other human beings? It's a brave new world I'm talking about here, and I for one can't wait for it to arrive.

No comments: