The future of…

Click here for contact details, other articles and personal details

The future of creativity

May 2001

One of the most common prejudices that we hold about machines is that they will always be incapable of creativity, they can only perform human-given processes on things that humans give them. People willingly accept that calculators can beat them at sums and that computers can win at chess (but only by fast number crunching - they don't really play as well of course!), but genuine intelligence and especially creativity is something that most people think will always be human. But computers are catching up whether we like it or not.

If you wrote a list ten years ago of things that you can do and computers can't, many of the items would now vanish from the list. Computers can play chess at grand master level, they can write high quality music, they can improvise jazz, they can produce works of art that can hang comfortably alongside human creations. They can come up with original ideas. They can write novels, create jokes, compose poetry. They can develop new mathematical theorems and prove them, invent and design new drugs. They can write software and design hardware circuits that are sometimes better than the best that humans can produce. In some fields they are much slower than people, and in some areas they have a long way to go to catch up with us on quality, but they are improving, and we aren't, so they will catch up. And they won't stop improving when they catch up. Before long they will outperform us in almost every area of mental activity. But still we will insist that they aren't really creative, it's just a trick of number crunching. We will credit the programmer, and when we are told that the computer generated its own code, we will credit the person who designed the code evolution environment. And if that was machine-designed too, we'll dismiss the achievement as trivial. Creativity is so precious to our egos that we will keep moving the goalposts until there is nowhere else to put them.

Eventually, computers will be much smarter than us. They will be so intellectually superior that only they will have the ability to fully appreciate their works of art. Humans looking at it might be unmoved in the same way as a fly listening to Mozart. But is it art? Who can say? The word is thoroughly devalued anyway. I read this morning that a scrunched up piece of paper has been sold to some twit for £2000, and is a contender for the Turner prize. It is very surprising that critics are so easily taken for a ride. If they really are that stupid, maybe next year I'll submit a half eaten piece of chewing gum or an empty crisp packet, or a hamster turned inside out. I can generate ideas of that quality faster than I can type them. It requires insignificant skill, but it is apparently still true that a fool and his money are easily parted. This standard of 'art' was passed many years ago by random number generators on pocket calculators. There is an infinite number of random everyday life articles or events to sample, and picking one using a digital pin from the web takes a few milliseconds. At a higher level of art, the sort of real art that you or I would hang on a wall, computers are now able to make pleasing pictures in certain styles, though they wouldn't compare yet to the best of kind. Give them time. Very soon you won't be able to tell human from machine generated art. In music and novels, we've already passed those points.

But creativity is more than art. Companies spend fortunes sending their staff of creativity courses, making them behave like children, creating special rooms where people can go to be creative. Removing the person from the routine certainly works sometimes. But why bother with all the effort and expense? In the near future we will be able to replicate this kind of experience synthetically using virtual environments, with a variety of creativity tools throwing new stimuli at us, bombarding every one of our senses. They will drag us on excursions into imaginary and fascinating worlds, forcing our brains to explode with ideas. They will seed brainstorms with previous ideas from the field from other teams, and cross link concepts in idea generation matrices. They will make pseudo-random mutations on ideas to stimulate lateral thinking. You will be much more creative in the future, but you won't be any more competitive. Everyone will have access to such tools, so we will live in an idea glut, unable to develop things anywhere near as fast as we can invent them. In fact, we already are in many fields. I've seen hundreds, maybe thousands of ideas for products that will make use of future mobile technology. Only a tiny proportion of them will ever be made, and consumers will only be able to afford to buy a few of those. Worse still, software already exists that can invent things all by itself. It can use evolution techniques to circumvent existing patents by finding alternative ways of doing the same thing. The commercial value of creativity is falling fast and soon it will be worthless, available on demand free of charge. It may be core to our egos, but creativity is as susceptible to automation as doing sums, it just takes a little more software development.