Site Meter

Monday, February 15, 2010

Computer science is the new physics?

I've been meeting a lot of people doing research in various topics in either computer science, machine learning, or various areas of somewhat related applied math topics. I was discussing pure vs. applied research with someone, and they were telling me that it was a very good time to be doing relatively abstract research in computer science. It was perceived that any kind of results, no matter how abstract can have practical benefit in a relatively short amount of time.

This reminds me of how I imagine it to have been for physics following the creation of the atomic bomb. The physicists were seen as miracle workers. Give these guys some money, and they will do magic with it. It seems to me like we are coming to the end of this. Physics has lost some its sway on the popular imagination. Observing the public perception of the LHC, for example, its hard not to see the last gasps of this former power. I don't think the physicists have no culpability in the impression that the LHC may create black holes to swallow up the world. Although not actively pushed, I think that it is also not actively discouraged. The problem with this, and I think the whole unification of everything via string theory falls along the same lines, is that no real result can live up to this hype. No matter what happens, it will be a let down to the public.

On the computer science front, I suppose there may be fruitful years of research ahead. What personally scares me about this is that the research is about our own imaginations, and not about the world. (But then, maybe this is just my bias, from not being very involved in it.) I do hope that this research doesn't get too far ahead of itself, and leave the world, and people behind. I suppose that physics has its control aspect as well.

A friend/colleague of mine forwarded me this link to these lectures by Hal Abelson and Gerald Jay Sussman about computer programming based around Lisp. One can only hope that some of this kind of spirit survives in the discipline of computer science. It is a joy of discovery and appreciation of simplicity and clarity that has a kindness and humanity behind it.

For me, I prefer to stick with the science. I prefer to try to understand what's already here, rather than create infrastructure to dramatically change things. And on the creation front, I prefer more modest modes of expression.

Looking back, I'm glad we have found quantum mechanics, and it is said that CERN produced the internet. I'm sorry that it has to be at the cost of a certain group of people presenting themselves as magicians.

A final comment I'd add, is that even though I phrased this in terms of "popular imagination", I think there is more to the story than this. I think that it is accurate to say that there isn't much chance that elementary particle physics will produce practical/technological benefits at this point. Certainly its possible, and one never knows the results of research. But it certainly doesn't appear that the Higgs Boson has any practical benefit, nor do the superparticles for that matter. And I don't think many particle physicists think this either. I think they just believe very strongly in reductionism, and this is the main argument as to why particle physics research should continue. I have to admit, I'm sympathetic to this line of reasoning. But I also think something has gone wrong with this approach- it has gone too far- and it will be very interesting to see what the effects of LHC research are, whatever the results!

No comments: