Friday, 2 March 2007

War-blogger’s retraction

Listening to:

El cuarto de Tula, a live performance from a Buena Vista Social Club sampler CD.

Retraction is always hard

This essay is an impressive piece of political recanting. One, because I totally agree with where the author ends up, and two, because such public admissions of error are so rare. Contrast with The Economist, which still claims that the decision to invade Iraq was right given the evidence available to the concerned governments (even if the information was bogus, and quite possibly concocted to please the people asking for it), and that invading Iraq went so wrong because of poor execution.

If they’d stuck to Afghanistan, where it does appear as if NATO intervention is improving things, how much more kindly we’d judge them.

Friday, 9 March 2007

Detecting human-level intelligence

Listening to:

I wants to stay here, from Cole Porter’s Porgy and Bess. Sung by Ella Fitzgerald.

Way Cool Factor 10, Mr. Sulu

I was dead-impressed with this cool test for detecting human-level intelligence, by getting people to tell the difference between photos of cats and dogs. This is something that machines can’t do very well, if at all, but something that a child can manage with high levels of accuracy from a young age.

And when the machine learning gurus get their systems up to the task of telling cats from dogs, perhaps they will founder on judging whether or not people are attractive. That’s the premise behind There are two obvious problems with this latter though: it seems likely that a fair component of beauty is culturally determined, and once you take that out, maybe the rest is determined by “simply” assessed numbers like amount of flesh on display coupled with symmetrical faces...