2009-12-16

Science in the subjunctive mood

"One difficulty in discussing the propagation of classicality through the Complete Feynman Diagram is that the term 'propagation' suggests that we are discussing a temporal phenomenon, which we are not."

2009-12-15

Connoisseurship is not important

No, see, connoisseurship is a hobby. It's not important. Well, except maybe insofar as it's good to hone your senses. But otherwise, no. Please do not think that I think that it's important.

2009-12-10

incomplete

so what does Gödel's Incompleteness Theorem actually give us as a take-home message? what's the upshot? we treat it with great reverence, but what does it actually mean? it's usually accompanied by some parenthetical words about its meaning, when it appears in works for the layman, but they're often quasi-mystical and never fully worked out, much less verified or even verifiable.

2009-12-07

Do you tweet?

I just now worried that I hadn't checked my email for a while and that there might be people expecting me to, and then I thought, well, hey, I used to compulsively glance at my system tray for the Outlook new-mail icon every half a minute; is it so bad that I've re-architected not to do that any more?

See, stuff like that doesn't fit in a tweet, so I gotta keep this blog.

2009-11-17

I know you know I know.

so if everyone had Logical Fallacies as part of their education, it wouldn't matter if everyone remembered the specifics. it would matter that everyone knew that there were fallacies, and that everyone knew that everyone else knew.

2009-09-24

Polling

When a computing process wants to keep informed about new information, one technique engineers use is a polling mechanism. Polling is when you go check in a known location to see if there's anything new waiting for you there. Engineers debate about setting polling intervals and when not to use a polling mechanism, and so on, because polling consumes resources. You have to be careful when designing a system to make sure it doesn't demand that a computer do too much polling.

But nobody seems to mind when you design a system that demands that the user do lots of polling. Your mailbox at home, your work inbox, your email inbox, your other email inbox, your other other email inbox, your Twitter page, your LiveJournal inbox, your facebook inbox, your voicemail, your cell voicemail, your RSS reader... all of these demand that you go check them periodically. Which is really not an efficient use of your processor. And every time somebody adds a cool new communication service to the world, they add one more place that everybody has to poll.

2009-09-01

Temptation

An angel offering eternal salvation to a man is like a hot teacher offering sex to a teenage boy. No matter how sure he is that he wants it, he really doesn't know enough to fully comprehend the scope of the offer.

2009-06-23

Net

Jesus, Christ, all these thoughts. A torrent.

Writing my thoughts is like dipping a goldfish net into a river. Too small goes through. Liquid goes through. Too big breaks the net. Too clever or quick wriggles out. Too far from the bank is entirely out of our purview. And what about before I arrived, and after I leave?

So if something solid and dumb that is bigger than the holes in the net and smaller than the whole net happens to pass close to the bank at precisely the moment I dip the net, I might catch it.

And why am I sitting on this bank with this net? What do I want out of this river?

If nothing I want fits those criteria, what's the point?

2009-06-04

The keliophores impede the crawler's motion. With our dexterity thus reduced, the chances of our missing a branch are increased. We may have to rest here until we purify more of the solvent, but the prospect of delay, any delay, terrifies us all.

2009-05-27

Corners

Who is standing in the space between Kurzweilian infatuation with the notion of machines with superhuman intelligence, and Baxterian pessimism about the notion of collectives with subhuman intelligence? Am I alone here?

2009-05-19

Sir? Step over here, please.

If a doctor (or inventor, or company, or whoever) invents a medical device to cure or treat some sort of sickness or disease or disorder, she can't use it, can't put it into practice, until she's proven that it does what she says it does.  Why aren't security professionals held to anywhere near that standard?  

I'm pretty sure stents for atherosclerosis weren't used on a patient in a hospital before somebody asserted to somebody else that those stents did something useful, and backed up that assertion with some kind of proof.  Did any similar process happen before airport security started demanding that we take off our shoes before we could board a plane?

2009-05-08

Bottleneck at the Retina

I really want dual monitors. Or a secondary mini-display. Or one really big display. Or a head-mounted display. Or an implant.

2009-04-29

Moot

When the end of the Bafflement comes, will we be ready?  

Ready?  There's no such thing.  

Well here it comes anyway.  

2009-04-25

All Greek To Me

When we learn the alphabet, we're presented with a big picture of a letter, in its uppercase and lowercase form, and shown that letter in use in a familiar word.

Nobody ever did the equivalent for me for rho, sigma, lambda, etc. They just showed up on a blackboard in a sophomore physics class, without introduction.

Is that a common experience?

If so, then physics educators are lacking basic techniques known to kindergarten teachers everywhere.

That's pathetic.

It's like hiding physics knowledge behind a secret door so only the most determined students actually reach it.