According to Kevin Drum, in a fascinating article for Mother Jones, artificial intelligence began in 1956. Or, rather, that’s when the term was first used. Ever since, enthusiasts have promised us that artificially intelligent machines are just around the corner. But thus far, they’ve yet to appear.
Drum has an explanation for the no-show:
However, thanks to the continuous improvement in microchip technology, we’re now within striking distance of making computers that can match the processing power of the human brain:
In other words, computers are getting more powerful at an exponential rate. Now, the thing about exponential increases is they have a habit of creeping up and surprising us. To illustrate the point, Drum devises the following thought experiment :
In fact, after the first half-century, the results would be barely detectable. By 2010, you’d begin to see some progress – a “few inches of water here and there” – enough for a paddle. However, filling the lake still seems like it will take forever:
Drum starts his thought experiment in 1940 because that’s when the first programmable computer was invented. And he chose Lake Michigan, because its volume in fluid ounces is the same as the processing power of the human brain in calculations per second.
Thus for every calculation per second that our computers could perform in 1940, they will, by 2025, (assuming Moore’s Law keeps going for another decade or so) be able perform 10,000,000,000,000,000 calculations per second.
That doesn’t mean that the robot rebellion is only twelve years away. As Drum admits, “raw speed isn’t everything”. There is, for instance, the small matter of consciousness – of which we’ve yet to see a glimmer in our mechanical pals.
What the exponential increase of computing power does mean, however, is that things that computers can already do to a limited extent – like operate driverless cars, understand human speech and interpret CCTV footage – they’ll soon be able to do to an unlimited extent.
This will have major consequences for public policy, but few politicians show much understanding of what’s in store. Right now we’re still at the paddling pool stage of the new era, but our basic assumptions over a wide range of policy areas – from transport to employment to civil liberties – are about to be swept away.