Wed, 09 Dec 2015 17:11:45 GMT
If virtual reality loomed over 2015 like the inescapable T. Rex in Jurassic Park, then the velociraptor sneaking up on 2016 looks set to be artificial intelligence. By next December we’re sure to be overwhelmed by ad agencies claiming to be AI experts, thought leaders musing about sentience and campaigns borrowing heavily from Isaac Asimov and Philip K. Dick. Tech giants like Facebook and Google and Microsoft are already investing millions in the area that brings together big data, the Internet of Things, computer science and neuroscience. And I’m not sure how I feel about it.
Somewhere between 11 and 13.5 billion emails flood my inbox every day, so the news that Google are set to launch an A.I. project that answers and organises your emails is definitely something I can get behind. It’s been created using ‘deep learning’, which uses algorithms and models from the world of neuroscience to allow machines to learn, react and adapt to information. And yes, the prospect of a constantly mutating programme scanning millions of emails might be a teensy bit of a privacy issue… but chronic email guilt wins out. And laziness, if I’m honest.
Another potentially handy email artificial intelligence programme is x.ai – or Amy, the AI personal assistant who schedules meetings for you. Brilliant – spare me the endless email ping pong.
But… having consumed far too much science fiction, I can’t shake off that little, niggling voice. The voice that says, “this never ends well”. Doesn’t no one remember Skynet? H.A.L? GLaDOS from the Portal games, the most vindictive of all the evil A.I.? And even when they’re not out to hurt you, they seriously bring you down. Marvin the Paranoid Android from The Hitchhiker’s Guide to the Galaxy shows just what you risk by imbuing machines with artificial emotions.
For someone who writes about creative technology and has interviewed some of the most innovative figures in the advertising industry, I’m an embarrassingly slow adopter. Natural paranoia combined with an overactive imagination and a desire that this ‘we live in an increasingly connected world’ nonsense would just go away will do that to a person.
The old cliché ‘we fear that which we do not understand’ has truth to it. And I have not a chance of ever understanding artificial intelligence. There’s a chapter in Steven Pinker’s book The Language Instinct that explains, step by step, how the Turing Test works. (The Turing Test was created by Alan Turing to test a machine’s ability to display intelligent behaviour that is equal to that of a human.) I tried to read it, I really did. I tried several times, in fact. And each time, my brain turned to mush after the first paragraph. Ten minutes would pass and I’d realise I’d spent it flipping pages and had no memory whatsoever of what they contained. In the end I just skipped ahead to the next chapter.
And there’s the problem… deep, deep down I’m afraid that this artificial intelligence will be more intelligent than me. We’re funny like that, human beings. Insecure little egotists. Sure a computer can do complex computations at a speed I’d never manage, but I can understand language, adapt to new circumstances, learn, create… If the boffins manage to crack sentience, well, I won’t even have that.
Our only hope is that A.I. scientists have been, as I have over the past week, binge-watching old Red Dwarf episodes on Netflix. I think I could handle an A.I. that was dopey, incompetent, slack and balding like Holly, the spaceship’s sentient operating system. In later series he becomes a vapid and indifferent Essex girl and I would also be quite happy with that too.
Look, what I’m trying to say is in 2016, let’s not get too carried away with the artificial intelligence hype train. Artificial stupidity is where it’s at.