Friday, December 31, 2004

e is for error correcting

random thoughts on new year’s eve. error correcting codes are cool.

I haven’t fully techie dorked out in a while (too much of a social scientist, sigh), so I thought I would explore this random part of my mathematics/computer science curriculum that comes up in weird random places: error correcting codes.

the basic idea is that an amazing number of parts of human life can be thought of as encoded message transmission: dna, language, music, art, puzzles. At mit, a large number of classes dealt with encoded message transmission but only recently have I started appreciating the broader applications.

a recent American scientist article talked about the efficiency of the dna-codon to amino acid coding. everyone knows that there 64 (corrected) possible codons (length 3, with 4 base pairs) but only 20 something amino acids. What’s cool is how amazingly efficient the encoding is, that allows functionality up to random errors in the signal.

This is the same idea behind how cell phones work to carry voice messages, how cd’s work to store music, how mp3s compress music, and many more. this train of thought was largely inspired by the adverts for dialup modem speed enhancers, which reminded me of the power of z-modem over Kermit and x-modem.

a code can be thought of as a subset of a message space. efficiently designed error correcting codes have an associated distance metric that places different code words as far apart as possible, so that many errors in transmission are necessary before you mistake one code word for another. the English language is such a code. i cna tpye a vrye grrbllde emssgge adn yru cyn ltlll rped it. economist ariel Rubenstein and others discuss the optimal design of languages.

a fun simple experiment is to run pkzip on a text file, and you’ll be able to demonstrate to yourself that English and in fact most languages has a message space 8 times larger than the code space. each letter in a paragraph typically conveys only about 1 bit of information.

other fun applications:

economist glen loury believes that the identities we use are codes to truthfully transmit information about ourselves, and then argues that the codes African Americans use are inefficient and self-defeating.

in analog signal processing, the nyquist frequency is a nice analog to what’s going on in this digial coding/encoding.

my favorite application is a riddle that I will end with. 10 prisoner’s are told they are to be executed. they will be lined up single file and each will be given a hat, black or red. They will be lined up in such a way that each prisoner can see the hats of each person in front of him, but not his own. The person in the back can see 9 hats, the person in front of him 8, the person in front can see none. From back to front, each prisoner will be asked to say either black or red. If he says the color of his own hat, he is saved, if not, he will be killed. Find the algorithm that in expectation, saves the most number of prisoners.

(Hint: 9.5 prisoner’s can be saved…)

Friday, December 24, 2004

d is for doping

Random weird connection. Some talk show on NPR talking about the recent sports doping scandals made the interesting connection between sports doping and other realms of performance enhancing drugs, i.e. academics and music. I never personally knew of anyone who was into sports enough to be a likely steroid user, so the prevalence of it though widespread didn’t really hit me.

But the expert suggested that there are college students who take un-prescribed Ritalin to help them study for exams, and musicians who take beta blockers to help them escape stage fright. I know people who do both.

Who knew that Ritalin, miracle treatment for ADD would work for anybody to provide added focus. And then you have to wonder, is it really wrong.

It was shocking how casually this one girl admitted to the Beta Blocker trick, for a high school violin recital of all things. These drugs prevent your heart rate from increasing. It is something I can relate to very well. I pretty much choke every audition I’ve ever been in, because my heart must triple in speed to a rate I never feel in any other situation, including exercise. But a drug seems like a ridiculous measure. The girl joking warned to be careful. She had gone jogging after the recital and passed out because her heart couldn’t provide enough oxygen since it couldn’t beat fast enough.

Monday, December 20, 2004

b is for bookstore

for a time, I never understood bookstores (at the time, these to me were lifeless interchangeable national chains) when one could have an infinitely wider selection online at amazon, until I realized that today, a bookstore is not valuable for how many books it carries, but how few. (see also my epinions note: )

editors are important to tell us what's important. Which is why record companies and tv networks will always have a job, even when content becomes essentially free. we need people to tell us what’s good.

when it comes to bookstores, a good one will tell you which few amongst the universe of millions are worth reading. and your favorite, will be the one that fits you, speaks to you. two have succeeded for me

thus, the best bookstores at doing that for me, is firstly the MIT press bookstore, a tiny bookstore across the street from the main campus bookstore, stocked to the brim with fascinating books related to all the knowledge MIT is interested in, from nanotech, to social networks, to typography, to urban design to poetry. the place is crammed with books that pique my geeky fancy.

recently, I rediscovered Kramer books in Dupont Circle in Washington DC. Aside from having a nice bar, live music and great desserts, this store reflects the social scientist in me. every book on display in there from hip snooty fiction like Chabon and Umberto Eco to pop intellectuals like Krugman, to general interest social science on topics like cultural social psychology, I’ve either read, want very much to read, and all I could bull shit about at length.

today I add a third for different reasons. Dewey’s Comics in Madison, NJ. A store I frequented a bit in high school about a decade ago, and liked especially because of how incredibly nice the owner was/is. in stark stark contrast to the jerk who owned the other local competitor Fat Moose. I tried to go back and now and then on visits home, and happening to be in Madison today, I stopped in. It’s been four years since my last visit (I know because I first went to their old location, which apparently they left 3.5 years ago). The owner was still there, and he still remembers me, though it’s been 9 years since I’ve been a regular customer. He even recalled seeing me in USA today. It felt like home.

Friday, December 10, 2004

divertissements - diversions

I have a lot of hobbies I noticed. At most of them, I am either the worst of those who are good, or the best of those who suck. Like I play clarinet for various Stanford musical theater productions. I'm typically the worst musician there, but passable at least. I still play a couple times a year, so I'm better than the great majority of those who played in high school and then gave it up.

At hockey, a bunch of us absolute beginners started playing a couple years ago. After a bunch of ringers started playing, I'm one of the only one of the original gang that still comes. So I'm better than all those beginners, but the worst one there now.

At chess, I'm good enough to know how bad I am, which is quite a feat. I know what a queen's gambit acepted is, and the importance of initiative, but that doesn't stop me from losing the initiative to Adam every time we play, and for him to whoop my butt every time.

Anyway, I only bring these up because last week, for these things. My mind was largely occupied with my own research (I gave a seminar at the end of it), and so I got crushed at chess, and poker, and had a shitty performance.

But this week, seminar over, brain unoccupied, I kicked ass at poker (won $45), did well against adam at chess (won like 2 out of 6 games, twice as many as I've ever won against him before), and had a kickass performance as the clarinetist for Gilbert and Sullivan's HMS pinafore.
ok, enough babbling.

Monday, November 29, 2004

The ownership economy and other columns

My columns for the stanford daily can be found at I am particularly proud of the first one. They forgot to post one of them so I have posted it here:

The ownership economy

I may be shunned by the economics profession for saying this but I am a closet Marxist. The academic economics profession in the US has been the only humanities/social science department to have effectively exiled their Marxists to peripheral schools. What is under appreciated, is that they have done so after adopting and fully internalizing what is in my opinion, Marx’s most important contribution, dialectical materialism.

Though Marx is often associated with notions of communist ideology, the ideological aspect of Marxism was introduced by his followers such as Gramsci, decades after Marx’s death. Marx himself, at least in his more academic and less rhetorical writing, was a strong believer that material factors – tangible things such as amount of labor or capital – determined the shape of the economy. This materialist approach is central to the mindset economists have today.

One other point of agreement between Marx and modern economists that Martin Carnoy likes to point out is that capitalism is the best system at wealth creation known to man. The crux of Marx’s argument against capitalism is not that communism would provide more wealth, only that capitalism distributes wealth unfairly. Capitalism concentrates wealth in the hands of the capitalists.

The question I always had, then, is why can’t everybody be a capitalist?

I was reminded of all this by the recent rhetoric of the Bush campaign, and his vision of an ownership economy. Bush wants to create programs that allow all Americans to have greater ownership of their homes, their healthcare savings and their retirement savings, i.e. their capital. I would argue that this is a Marxist idea.

One simple point that is taken for granted by economists and their models of the economy but is easily forgotten or ignored by everyone else is that corporations are owned by people. People seem to have a visceral distrust of corporate profits. However, corporate profits eventually go to the owners of the corporation, i.e. the shareholders, or as Marx would call them, the capitalists.

In America today, over 50% of Americans own stocks. That makes them capitalists. I see no reason why this number can’t increase. Admittedly, Marx and many others would point out that ownership of capital is very unequally distributed, and that most of these are petty bourgeois at best, fooled into thinking they have power, but just as oppressed as the rest.

Yet that could change as well. The main reason Marx’s predictions did not come to pass—he expected England to be the first communist country—is that workers started acquiring skills and education, i.e. human capital, that allowed them to start demanding higher wages. Though estimates of human capital are imprecise at best, many estimate that one third or more of any nation’s assets are in the heads of its people. Marx’s original theory was not wrong, only blind to technological developments.

By improving education, and transferring ownership of assets from the government to the people, we potentially move toward a society where everyone’s a capitalist and everyone’s equal.

I further want to disabuse people of the notion that capitalism requires inequality in order to produce incentives. Untrue. Capitalism requires only the possibility of inequality. In a society where capital is evenly distributed and everyone works equally hard, there is no economic reason why capitalism would not produce an equal society.

John Locke and the founding fathers held the belief that ownership of property was an essential component of citizenship. I am inclined to agree.

Thursday, November 18, 2004

Clinton and Me

In today's Nytimes:

Bill Clinton:

"Am I the only person in the entire United States of America who likes both George W. Bush and John Kerry, who believes they're both good people, who believes they both love our country and they just see the world differently?''

Wow, Clinton stealing a line I wrote almost word for word. Scary.

Saturday, November 13, 2004

geo-political prognostications 2004

So I actually predicted over a year ago that the economic recovery would kick in and the war in Iraq would be settled enough, such that Bush would win re-election. Unfortunately, I never wrote it down as proof.

So, now I may predictions for the next four years. People tend to see the war in Iraq too simplistically. Because victory was not easy, now they assume it is unwinnable. My feeling is that the truth is in the middle. Yes, post-war Iraq has been a mess, but it is not unwinnable.

My prediction is that Iraq will resolve much nicer than expected, the elections in January will be the beginning of stability, and while a shining example of democracy will not arise, something resembling democracy will take hold, providing a shred of hope in the middle east.

This will be strengthened by the Israel-Palestine conflict. The choice of Palestinian succession will be easier than expected, and the new leader will help take the first steps that finally resolves the generations long conflict.

These two successes will leave Bush’s legacy in the middle something that will be remembered in the history books.

Sunday, November 07, 2004

The GI and the Chocalate Bar

In a recent NYTimes artlce, there is the following quote:

"he developed a love for all things American when tanks rolled into Paris and G.I.'s rained chocolate and chewing gum on children like himself."

I always loved that image. In one of Stephenson's books (i think), an old Japanese executive is deeply loyal to the US, because as a child during WWII, after being told of the monstrosities he is told to expect, his first encounter with a GI is with a chocalate bar. The other similar image is the end of Life is Beautiful (one of my favorite movies), as the boy with such a beaming smile is pulled up onto the tank...

I suppose it was fantasies about this image that led to misexpectations in Iraq, but it is a nice image nevertheless.

Wednesday, November 03, 2004

Why Democracy Sucks

An uncut repost of my column from the Stanford Daily, also available at

(For the record, I called Bush's victory over a year ago. I thought I had put in on my website somewhere, but appraently I forgot. The economy would recover, Iraq would be indeterminate and Bush would win. Not that this means anything.)

Why Democracy Sucks.

In just three days, Americans, at least half of them anyway, are going to exercise their inalienable right to vote, so I thought I would take this time to assert that Americans are too stupid to decide who gets to be president. Actually, not so much stupid as ignorant. I freely place myself in that category.

The useless presidential debates consisted of two politicians repeating oft-heard sound bites, containing little substance. But who can blame them. My friends watching were mostly commenting about facial tics, vocal inflection, and accidental misspeaks. No one cared about policy questions or substantive issues. Who could be bothered?

I was further disheartened listening to NPR’s coverage of California’s ballot propositions. I was horrified by how ill-informed the experts were, much more so the callers, and could only assume the listeners were worse. As for the vast majority of voters who don’t even pay this much attention, I shudder.

I find it disturbing how people who devote just a few minutes of their days to these issues feel qualified to second-guess policy makers who devote their lives. This is exacerbated by campaigns such as Rock the Vote which encourage the attitude that your opinion, no matter how uninformed, should count.

Two weeks ago, Jon Stewart self-righteously showed up on CNN’s Crossfire, lambasting the show’s hosts for reducing complicated issues to partisan spin. Yet CNN, like Stewart, is merely providing what viewers want. No one watches PBS’s News Hour.

Yet I believe the people too are blameless. I used to think education was the answer to all of today’s pressing problems. Like Bush, the answer I used to have for poverty, inequality, affirmative action, democracy, etc., was better education.

Recently, somebody asked me which candidate I favored. I equivocated because incredible as it may seem, I honestly like both. Yet she pressed: if the election were a tie how would I vote? I admitted I don’t know. I am not nearly qualified to make such a decision. I—product of arguably the best education in the world—was insufficiently informed to answer.

The dirty little secret of rational choice political theory is that it is basically irrational for people to vote. The chance of one vote affecting the outcome of an election is so miniscule that it makes no sense for people to bother, much less expend the effort to properly learn the issues to make an educated decision. We’ve known this since Plato.

Philosopher John Dewey had a beautiful thought though. He argued that since musicians and poets can convey incredibly complex ideas to a broad audience, social scientists would one day learn to convey complex policy questions in a way that could be broadly understood. However, unless the Daily Show is the beginning of a revolution in social science communication (just crazy enough to be true), we are a long ways off.

So do I have a solution? Well no, though I’d first abolish the Californian proposition system, pernicious legacy of the progressive movement. However, I might argue that maybe the spittle and facial tics during the debate really matter. Maybe since voters are ill-equipped to judge the policies of the candidates, the best they can do is to assess their character.

Better still, I propose a radical new system of elections with a rule that you can only vote for somebody you’ve personally met. Hanging out in Singapore, I was surprised to find that despite the essential absence of political competition, Singapore seems reasonably well governed. The secret, a senior minister told me, is that members of parliament are able to meet their constituents personally. He holds open houses every Wednesday, and spends one day every weekend going door to door.

The personal contact rule takes advantage of the millions of years of evolution that have left humans with social and cognitive tools extremely well adapted to interpersonal interactions. Judgments are aided by networks of social connections. Candidates should be friends or friends of friends. Bad candidates are quickly identified and removed from office. Then, the candidates, one for each community, collectively choose a new executive. Essentially, the electoral college as originally envisioned.

Of course this system has many problems and is mostly just a pipe dream. But it’s a thought.

Not that it even matters. Psychologists know well that we attribute too much credit to the person, and not enough to the situation. The character traits of the person in office matters a lot less than we think. In a recent cross country analysis, two economists, Jones and Olken, find that the policies of democratically elected leaders have no detectable impact on economic growth. Incidentally, they do find that the policies of autocratic leaders significantly retard development.

Monday, November 01, 2004

New Look: Blogger Trial

After four years or so of maintaining a web log long before the word blog was invented, writing my own tools in perl, it is time to try something new. I have to admit that I am no longer techie enough to do it myself. The time's they are a changing.

The old posts can still be found at