Monday, February 15, 2010

DIY genetic engineering

Here is a link to an Interesting Article from Sunday's (2/14/10) NYT Magazine: Do-It-Yourself Genetic Engineering. The article describes the bubbling new field of synthetic biology, a a step or two beyond traditional genetic engineering -- instead of adding new genes to existing species to change their properties, synthetic biology is more about assembling new organisms from genetic legos.

It's more next-gen technology revolution stuff: "synthetic biologists imagine nature as a manufacturing platform: all living things are just crates of genetic cogs; we should be able to spill all those cogs out on the floor and rig them into whatever new machinery we want." With the added possibility of creating self-replicating factories.

Which is both fascinating and terrifying -- the article references "bio-terrorism", but that's only a minor issue I think compared to what the article describes as "bio-errorism". When fiddling with stuff, if anything can go wrong it will -- that's just the nature of technology. Mainly you want to minimize downside risk so that no one gets hurt. Nature-as-genetic-laboratory is incredibly cautious. It has a vigorous quality assurance department, and runs any new developments through a lengthy testing and shakeout process.

[Cynicism / realism alert] Why am I not so confident about the same when money is on the table?

And then there is the utter reduction of nature to machine. Is that all there is?


Saturday, February 06, 2010

Fresh Air interview with Ed Thorp

The Fresh Air radio program recently featured Ed Thorp, mathematician and author of the classic card-counting book, Beat the Dealer, and Scott Patterson, Wall Street Journal reporter and author of the new book, The Quants: How a New Breed of Math Whizzes Conquered Wall Street and Nearly Destroyed It.

From the interview, Patterson's book sounds like a re-hash of how mathematical models, supported by computers, have changed the financial system. New technologies have created the possibility of new forms of speculation, and the demand for maximum rates of return has taken advantage to that possibility.

Thorp comes across as dispassionate and insightful. He recognizes the proper role of models in general. It's not that mathematical models of the world are bad, they just have limits. The "quants" (those who rely on math models for trading, as opposed to, say, the fundamentalists who are studying balance sheets) on Wall Street today "don't have a long history of experiences of what's wrong with models." They are "wowed by the math" and "didn't use them correctly." And so recent history is likely to repeat itself, somehow. "You know something bad is going to happen, you just don't know when."

Although it is popular to call Wall Street (or the modern financial system in general) a giant casino (which Thorp does in the interview), it is important to note that someone like Thorp was not a gambler, even when he was confirming the validity of his models at the blackjack tables in Las Vegas. On Wall Street, he did the same thing. Gambling and the application of theoretical models to acting on probable outcomes of current events share the important factor of risk, but they are fundamentally two different activities.