Wednesday, April 05, 2006

I was at the gym a few minutes ago...

(... I now look like this)

Huffin' and puffin' I bench-pressed 95 lbs, and the guy after me (whom I spotted) did 150 no sweat. Still got a ways to go...

... we got to talking, and it turns out the man does theoretical computer science, and I'd actually seen his advisor (Santosh Vempala) talk about some of his work. Good stuff, and Santosh is a good speaker, so I'd enjoyed those talks. Quirky, but that's de jure for CS theory (much more so than in the significantly stuffier environs I inhabit), and the enthusiasm about the work shone through it all and made it worth watching. That's the thing about mathematicians -- they know they're not getting paid to spin or be useful -- they're getting paid to look pretty. Or rather, do pretty... research.

Anyway, this reminded me of a talk I attended this afternoon. This was a crypto seminar; I had been warned about the crypto seminars by Sudan (the one with the Sudan algorithm -- his name's Madhusudan...) -- he'd told me that they really dig in at these talks, ask questions like it's nobody's business, actually understand stuff; you have to be a man to give a talk there.

I saw what he meant -- the speaker, an extremely good sport, on top of his game, a man with a brilliant Colgate smile, and obviously expecting this grilling, spent more time fielding questions throughout the talk than actually speaking. Related problems were explored, sharp questions were asked, naïve questions were asked, bad jokes and multi-lingual puns were made, tongue-in-cheek slides a la Seinfeld were shown, people laughed, and in general people had a good time. It was fun.

This is the world I've been entering, bit by bit, these last eleven years. It's very different from the external, non-scientific world's image of theoretical research. It's populated by a wide spectrum of people -- sure, there's some people who're... socially challenged reclusive eggheads, but there're also some of the most charismatic, knowledgeable and finger-in-every-pie sort of people I've ever met. It's also a lot more social an activity than people imagine -- much more research happens between people than happens in a single mind -- "Whereof one cannot speak, thereof one must be silent."

I also couldn't help but contrast this talk with one I attended yesterday by "the thinking man's Tom Friedman" (which is how he was introduced by the host). A sad underutilization of my time -- he said nothing really new, had no real insights. But in comparison, the most significant difference was really the one that always exists between talks in soft sciences and harder sciences -- in the latter, if you understand it, you walk away with a replicable, undisputable, truth; not a subjective opinion, but a honest-to-god absolute truth. The truth can be "useless" in almost every sense of the word (most research is, and if I'm honest, much of my research is), but that's not the point -- it needs to be judged by it's own internal standards of beauty (and in my humble opinion I've been lucky enough to stumble across some really pretty results in my research).

Now you, reading this, gentle reader, have either experienced the numinous when you understood a beautiful truth, or you haven't. And it's sad, but true, that most people haven't experienced the beauty of a fundamental truth that arises from internally consistent logic (rather than having spiritual roots; I'll grant you -- those moments can probably feel similar).
Which is kind of frustrating, because so many people are missing out on so much.

Some popular science books, shows, and movies do try to tackle this and communicate the camaraderie of people who commune with truthful ideas -- Proof, Numb3rs, Contact... but
even the best of these can't help but sortof skip over the hard parts, the parts where the essence of the matter, the ideas, are involved. They deal more with the interactions between the people doing science than with the science itself. Quite understandable -- science is unforatunately a chore to people, and has little place in entertainment.

Other works -- the thinking man's science types... Cosmos, Gödel, Escher, Bach... are enjoyable reading, but they have two flaws -- however much they profess otherwise, they already aim at an audience which is already familiar with the material to some extent, and also, they often miss out on the human aspect of doing research. By the human aspect I mean more than just the historical treatment of a few stars -- I mean the environment in which ideas are generated. I mean something along the lines of that talk I attended today.

And now, ladies and gentlemen, the highlight of this post. Here's my solution -- a fictional account of a graduate stude... an apprentice alchemist working on the advanced theory... repeated additions (i.e., multiplication), at the Peloponnesian Institute of Technology and Science. He's worried about where his next free dolmades will come from. He has deep conversations about his work with his hovel-mate, who's working on repeated subtractions. He's afraid of the visiting scholar from the QUranic Institute of Technology, this venerable old sadist with the mind like a scimitar and a tongue to match, but whom his master has tasked him to speak to and learn from...

I could go on and on, but the idea here is to base everything around a problem which is simple enough (multiplication) that everyone already understands it to a reasonable level of approximation, and can follow the convoluted process of many bumbling fools blundering around discovering ever more efficient ways of doing it, with ever greater insights into related problems and resulting in a finished, polished product that school-kids can learn by rote. This is what research feels like, and even the best and cleanest ideas by the smartest people (say for example Newton's Law of Gravitation) have hidden nooks and crannies that their discoverers don't realize, have a cleaner exposition, a proof, a formulation, implications for other theories, all in all, an iterative approach that only infinitestimally slowly converges to a place in The Book.


Blogger Vijay D said...

Hey Jaggi. Checking the blog after a long time. If you haven't already, you should try Simon Singh for the human portrayal of scientists, for capturing the excitement of the research and for telling a story about it all. His book on Fermat's Last Theorem was enjoyable reading for me. I still have to try The Code Book.

4:35 PM  

Post a Comment

<< Home