The singularity and simulation

Ray Kurzweil has a new book coming out so thereve been a lot of interviews with him. Odd that he never seems to mention simulation, when it is so relevant to his case.

Kurzweil argues (eg in a recent New Scientist interview) that the rate of change or growth in human society is accelerating expnentially.(“Between 2000 and 2014 well make ..progress equivalent ot ht entire 20th century..”) Information technology in partcular will grow at an explosive rate. This will lead to a singularity, when change will be so rapid and deep that human life will be irreversibly transformed. Kurzweil says: “Were already in the early stages of augmenting and replacing each of our organs, even portions of our brains…hundreds of applications of narrow AI – machine intelligence that equals or exceeds human intelligence for specific tasks – already permeate our modern infrastructure…. ultimately, we will merge with our technology… keep in mind that non-biological intelligence is doubling in capability every year, wheareas our biologial intelligence is essentially fixed.”

The Economist Technology Quarterly quotes a good example: software programmes that buy and sell shares and forex are outperforming humans. Not only are the programmes faster, they can also take far more variables into account. Its a logical development of chess-playing programmes: at first we thought it wouldnt be possible for a machine to beat the best humans, now its fairly routine. But chess “doesnt matter”, whilst forex trading does. People go hungry if someone (or some thing) gets it wrong.

An article by Robert Axelrod, “Advancing the art of simulation in the social sciences”, argues that simulation can be used for prediction, proof and discovery in the hard sciences, and also that in social sciences it can “discover new relationships and principles”. Axelrod sees its main contribution as the sutdy of adaptive behaviour. Where responses can be rationally deduced (eg the assumptions of neo-classical economics) we can sit down and work out how a situation will develop. But where actors adapt, it is impossible to deduce the consequences. Simulation cannot predict, but can show how situations may develop. (I suppose swarm theory is a good example of a new viewpoint emerging from this new tool.) Certainly the huge amounts of simulation power now available to almost anybody should massively increase our ability to understand adaptive behaviour: and adaptive behaviour is what we all do, no matter what Samuelsons “Economics” (etc) used to tell us. So we ought to be able to discover new relationships and principles at an ever-growing rate as well.

Incidentally Im unhappy about Kurzweils use of the word singularity. In physics it implies an end to life as we know it (eg a black hole); in mathematics it can imply zero or infinity. Its a good catch-phrase, but I think what it really means is a paradigm change. I suppose much the same thing happened with the invention of the telegraph (communication unbundled from transport) or printing (ie knowledge unbundled from unique artifacts). Kurzweil would probably disagree. However, Let me end with another quote from him:

Leave a Reply

Your email address will not be published. Required fields are marked *