I recently read a prediction by Ray Kurzweil that made me stop and think.
Kurzweil is Google’s ‘leading futurist’ and apparently a man greatly respected by Bill Gates. The prediction in question was that human brains will be directly connected to the internet, using nanobots.
Now, leaving aside the nanobots for a moment, this is a prediction I’ve got a lot of time for. It seems clear that not only clunky old keyboards and mouses, but also touch screens and ear-phones can’t be long for this world. Generally speaking, things are a damned nuisance, and carrying around a device of any sort is a bad idea. The smaller it gets, the easier it is to lose, to sit on or to have stolen. So why not internalise it?
It doesn’t have to be a big deal when size isn’t the issue. Just put a tiddly thing under the skin next to the skull. No wires, you can train yourself to control it using brainwaves, and (perhaps more of a challenge), to ‘see’ images and ‘hear’ text from it. So, you can surf the internet any time you want, get any information you need, watch a film or play a game in full 3D just by closing your eyes. You can communicate with anyone who has a similar device without having to open your mouth.
What we’re talking about will appear to be telepathy, almost. Will we need to speak any more? Most of the technology is already there – controlling computers with brainwaves is the nub of the concept, and that’s been done.
Nice idea? Well I guess most people will have immediate reservations (futurists should always bear in mind that folk are largely conservative and suspicious of change). The consequences would certainly be far-reaching, hard to guess at and not all positive.
Your ‘computer companion’ would help you through all the mundane details of life; a little voice reminding, advising and consoling, learning to be perfectly in tune with all your needs and idiosyncracies. You would quickly become dependent on it – quite lost without it, in fact. The most important point, though, is not whether or not it’s a good idea, but that it’s ‘cool’ and it makes life easier. For those reasons alone, people, conservative as they may be, are likely to buy into it.
As we talked about this idea, someone I know mentioned Philip Pullman’s ‘His Dark Materials’. It struck an immediate chord with me. In case you don’t know the books, in Pullman’s universe everybody has a ‘daemon’, a sort of animal companion that lives closely with you throughout your life. This entity helps you, reacts to your thoughts and emotions, sometimes argues with you and will try to protect you to the last fibre of its being. To be separated from your daemon is a terrible and painful trauma for both parties.
It’s a compelling thought – could your ‘computer companion’ become a sort of ‘daemon’? A sweet, wise and supportive friend? The greatest love of your life? Would it report back to the authorities any ‘inappropriate’ thoughts you might have?
But what’s Ray Kurzweil’s take on this? To start with, the nanobots connecting us directly with the internet will be made of DNA: “They’ll go into our brain through the capillaries and basically connect our neocortex to a synthetic neocortex in the cloud”. In fact, we will become ‘cyborgs’. This will happen by 2030.
Hmm. These nanobots seem interesting, but I’ll bet there’s nobody alive today who has a practical idea about how to create them – better hurry up, there’s less than 15 years to go. Not much about the sort of world these ‘cyborgs’ would create and inhabit, either. Kurzweil nods in the direction of the risks: “Fire kept us warm and cooked our food but also burnt down our houses. Every technology has had its promise and peril”, but on the whole he doesn’t seem to have thought much about the social implications. He doesn’t seem that interested – it’s technically exciting, and that’s enough.
I had a look at some of Kurzweil’s other predictions (by his own estimate, he’s right 86% of the time). There was the one about a computer beating a human at chess by 2009; he also predicted that computer displays would be built into glasses and that by the 2010’s it would be possible to translate spoken sentences into a foreign language text.
And out into the future? By the 2040’s, non-biological intelligence will be a billion times more capable than biological intelligence (yep, a billion). Also, nanotech foglets will be able to make food out of thin air and create any object in the physical world “at a whim”. I have to admit I don’t have any idea what foglets are, but they sound kind of impressive.
He does admit to a few failed predictions – the self-drive car which we would be running about in by 2009, for example. But four experimental vans were tested successfully in 2010, so he sees this as more of a hit than a miss.
What are we to make of all this? Firstly, most of Kurzweil’s predictions that have come to pass are obvious – a computer capable of beating humans at chess has been on the cards for decades. In fact, it’s a tribute to the subtlety of the human mind that it took so long. Chess is a game that could have been designed for computers to win at. They win by crunching numbers. End of story.
Other predictions are technical ones, straight from Kurzweil’s area of expertise. Google did indeed bring out their (not very successful) computer glasses. You can’t escape the suspicion that Kurzweil was working on the early stages of the project when he made the prediction.
Then there are predictions that are, in any meaningful sense, just plain wrong. Yes, computer translators exist, but anyone who has ever used one knows how dumb and literal and often downright puzzling their output is. You’d have to be crazy to trust one for more than a very simple sentence.
And the longer term predictions? Well, I admit the nanotech foglets sound fun… but in a couple of decades? seriously?
I have to say I was rather disappointed by all this. Why had the interesting idea of a human/computer interface been over-complicated and extrapolated into wild areas? Here was a man in his sixties, spoken of as some spookily accurate predictor of what’s to come, addressing the future like a bouncy twenty-something who thinks IT and the Internet are ‘really cool’. Well, they are really cool, but can we perhaps calm down and then broaden our thinking a little?
I don’t want to make this into an attack on Ray Kurzweil, who I’m sure is a gifted innovator and ideas-man there at Google. Some of his specific predictions within his industry are doubtless very accurate. He may come across as something of a self-publicist but, hey, if you don’t blow your own trumpet, nobody else will… except for your good friend Bill Gates, perhaps.
There is an important point to be made, however. Kurzweil is a big believer in Moore’s law; the idea that computer power will double every 18 months. That’s exponential growth with a vengeance and remarkable for the fact that it’s exactly what has happened ever since Gordon E. Moore came up with the prediction in 1965. You can work out how many doublings we’ve had so far.
But will it continue into the foreseeable future? Kurzweil thinks it will do just that– or even accelerate if his ‘billion times more in intelligent than us’ computers are to arrive as promised in the 2040’s. It’s worth remembering that our most intelligent computer to date could be out-thunk by a sea cucumber.
Extrapolation. Surely that’s where most predictions fall over. Want to know the future? Look at the up-to-date things that are happening now, then – extrapolate. Just follow that trend, wherever it leads, and… wrong. In the early 20th century, a big worry for London was accommodating and feeding the growing number of horses on the streets– and dealing with the consequent manure. The motor car had been invented by this time, but it was just a rich man’s toy, after all.
Here is a wonderful magazine page from the 1950’s, illustrating what life would be like around the year 2000. We would eat food made from sawdust, fly to work in an aero-car, shop by television-phone and clean our water-proof living-rooms with a hose. All the ideas are amusing, most totally pointless or impractical.
But another thought struck me; if you notice, it’s a man flying to work in his aero-car while it’s his wife who buys a pretty dress over the television-phone and sprays down the house before putting the sawdust on the solar range in time for hubby’s evening meal. That’s right – she’s the little 1950’s housewife, still busying herself at home half a century on! By the 1950’s, in the aftermath of two world wars, the role of women was well on the way to change. It was quite obvious to anyone with eyes to see, but that massive social movement went straight over the heads of the ‘futurists’ of the time.
If extrapolation is one occupational hazard for ‘futurists’, so too is missing the (in retrospect) blindingly obvious. Maybe social revolutions are even more difficult to see, as they come whizzing out of left field, than the next breakthrough invention. And the one often feeds off the other, adding another twist of delicious unpredictability.
Perhaps Gordon E. Moore could be described as a remarkable ‘futurist’. After all, he got it right, didn’t he? But in fact, Moore thought the doubling of computer power would continue to around 1980, before slowing. It was only when the doubling continued (and continued) that the law was carved into stone. What Moore did was make an accurate prediction of the short-term future, but he was too cautious (perhaps too realistic) a man to go further. He should have extrapolated wildly, but… who can see into the future?
So can predicting, at least beyond the very immediate future, be anything other than a wild goose chase? We’re stuck in our own time and that means that all our points of reference are to the past. What can we do other than take what we know and put a little spin on it?
I think we can do a better job if we are at least aware of the pitfalls. Extrapolating the present into the future can even work for us if we do it with care. We should remember that things don’t usually happen (outside dystopian sci-fi) if people don’t want them to. We don’t eat sawdust because even if it could be made digestible, it would still taste bad, nor do we want to suck our dinner from a plastic tube just because the Apollo astronauts had to. I’ll have the roast chicken with fries and peas, thanks very much – oh, and give me a knife and fork while you’re about it. I’m going to put framed paintings on my walls and listen to classical music. If I can’t have a garden, I want a park where I can sit under a tree and read. Just because something can be done, doesn’t mean people will want it. If they don’t want it, there’s little incentive to create it.
Add to that a little common sense. Where would you park your aero-car? What would an aero-traffic jam be like? Why plant trees to get sawdust in 20 years when you could use the land to get carrots in a few weeks?
And missing the blindingly obvious sign-posts to the future that are all around us? We’re real champions at that. I’m talking, of course, of the dreaded ‘Black Swan’.
But more of that creature later.