Tough luck geeks, we’re still smarter than silicon
For a smart guy, Steve Wozniak — the man who, with Steve Jobs, co-founded Apple — has some pretty dumb ideas. Speaking at a business meeting on the Gold Coast recently, Wozniak claimed that machines are becoming more intelligent than humans.
Wozniak was reported to have said ‘We’re already creating the superior beings, I think we lost the battle to the machines long ago. We’re going to become the pets, the dogs of the house.”
In Wozniak’s eyes, humans are going to become mere spectators to the doings of machines. ‘Every time we create new technology we’re creating stuff to do the work we used to do and we’re making ourselves less meaningful, less relevant’, he said.
In nerd-speak, this is called ‘the singularity’: a threshold when computer processing power becomes so powerful that it surpasses the processing power of the human brain. At that point, we have artificial intelligence where machines think, act and feel like humans.
No doubt this seems reasonable to people who spend much of their lives with their heads stuck in the abstract world of high-end computing. But to the rest of us, the idea of the singularity is just plain silly. Not only does it show an astounding ignorance about what intelligence is, but also a radical — and potentially dangerous — contempt for humanity.
Claims about the rise of intelligent machines have a long and inglorious history. As philosopher Hubert Dreyfus documents in his 1992 book What Computers Still Can’t Do, in the 1960s artificial intelligence experts were confidently predicting intelligent machines by the end of that decade. When we missed that deadline, our relegation to the status of Lassie was pushed back to the 70s.
As the years came and each prediction went unfulfilled, the timeframes were pushed back further and further into the future.
Alternatively, artificial intelligence experts revised what they meant by an intelligent machine. Instead of a computer with the capacity for language, for instance, they settled instead for teaching a machine to play chess. And, when a computer did finally manage to beat a chess Grand Master, as IBM’s Deep Blue did in 1997, beating Garry Kasparov, it seemed that the age of intelligent machines had finally arrived.
The problem, though, is that describing Deep Blue as intelligent debases what most people mean by intelligence. The fact that Kasparov could walk out of the room after the chess and game and pour himself a cup of tea — in addition to playing a mean game of chess — ought to show us that there is more intelligence in a single human hand gesture than all the computer chips in the world.
In other words, intelligence isn’t reducible to processing power as the Wozniaks of the world would like to think. While doing one thing very well is nice, intelligence is having the capacity to do lots of things across multiple contexts. The only way to describe Deep Blue intelligent is to first dumb-down intelligence.
Wacky ideas about intelligent machines might not be a problem if they were confined to a bunch of Silicon Valley programmers who really ought get out more. But, increasingly, the running down of intelligence to artificially inflate how intelligent machines are is becoming commonplace — and it affects us all.
As Jaron Lanier writes in his recent book You Are Not A Gadget, ‘People degrade themselves in order to make machines seem smart all the time. Before the [global financial] crash, bankers believed in supposedly intelligent algorithms that could calculate credit risk before making bad loans. We ask teachers to teach to standardized tests so a student will look good to an algorithm.’
Lest this be dismissed as the rantings of someone who thinks computers are the devil’s handiwork, it ought to be noted that Lanier is no slouch when it comes to technology. He’s described as ‘the father of virtual reality technology’ and he rubs shoulders with the elite of Silicon Valley.
Lanier’s examples could be expanded. Talk about precision-guided ‘smart bombs’ and unmanned drones washes over us, as if machines have magically minimised civilian deaths in war. It all sounds very clear-cut and clinical, but news of innocent men, women and children being killed because of incorrect information supplied by unmanned predator drones in Afghanistan tell a different story.
We have been able to build machines that create the impression of awareness, but calling them ‘intelligent’ is pretty dumb. If we’re to avoid a future in which we hand over political, economic, cultural questions, with all the moral and ethical challenges these bring, we need to stop abasing ourselves before technology and re-assert the primacy of the human.
Christopher Scanlon teaches journalism at La Trobe University and is a co-founder of upstart.net.au.
Read all about it
Up to the minute Twitter chatter
The latest and greatest
Good morning Punchers. After four years of excellent fun and great conversation, this is the final post…
I have had some close calls, one that involved what looked to me like an AK47 pointed my way, followed…
In a world in which there are still people who subscribe to the vile notion that certain victims of sexual…