•29 March 2010 • 2 Comments

Why don’t mad scientists run the world? After watching Neil Patrick Harris and reading Austin Grossman, you may be wondering the exact same thing. Scientists discover the secrets of nature, they are the explorers of ideas and the guardians of knowledge. Surely, they could have withheld the atomic bomb from the rest of mankind, or terrorized the governments of the world into submission with giant robots. Why didn’t Fermi and Einstein and Bohr form a secret cabal and hold the globe for ransom? That would have solved the problem of applying for government grants every few years to fund their grad student and publishing addiction.

But sadly, scientists are relegated to the sidelines, like Szilard wringing his hands and fretting right before the end of World War II, and Oppenheimer’s belated remorse. Or their efforts are frustrated by their backward times. In Scott Berkun’s excellent book, The Myths of Innovation, he gives many historical examples of scientists trying to end violent conflict through inventions which ironically led to increasingly destructive war machines. The Wright Brothers originally pitched the airplane to world governments as the final instrument of world peace: with enough long-range reconnaissance, democratic governments would supposedly be able to detect and neutralize any enemy movements well in advance. However, both the Hiroshima/Nagasaki bombings and 9/11 were due to aviation. To fix up the Wright Brothers’ mistake, Tesla dreamed up a peace beam that would shoot down enemy planes from afar, which of course no nation could possibly use as aggressors (face firmly planted in palm, Nikola…) You could argue that the Wright Brothers and Tesla were not really scientists, but in their inability to predict human nature, I grant them honorary status.

The well-known pacifism of academics definitely plays a role in the dearth of mad scientist supervillains currently at large. But I like to consider another long-term trend. This trend has become so ingrained in the pursuit of modern knowledge that it seems inseparable from science: specialism. Specialism is the idea that in order to make progress in any enterprise, one must concentrate on smaller and more sharply-defined categories of ideas and phenomena, ignoring and abstracting away other details as distracting and irrelevant. This is the idea behind division of labor, and the reason why we have a dizzying array of departments in most universities, scattered across disciplines and speaking incompatible jargon, as if punished for building the Tower of Babel (which I imagine looked a lot like a silo). Departments of Medicinal Chemistry which are separate from Pharmacy which is not the same as Pharmacology which is different still from Genome Science which is yet distinct from Molecular Biology. Or closer to my actual home, Human Centered Design & Engineering ≠ Information School ≠ Human-Computer Interaction.

The rhetoric of science claims only to follow the Scientific Method, which Francis Bacon brought down to us from heaven engraved on stone tablets. There is no mention of how or whether to divide up fields, this is a detail of management and human resources that has emerged as a consequence of our flawed implementation. But just as surely as the U.S. Constitution implies a two-party duopoly, science begets specialism.

Now don’t get me wrong, specialism has given us a lot of great things, but all of history is a paean to specialism. What about the limits of specialism? Limit the first, Not My Jurisdiction. (This is related to the phenomenon of Not Invented Here). Let us say that these slices of delicious citrus fruits represent our current domains of knowledge, our university departments, our “division of research labor,” such that machine learning experts do not have to worry about psychiatrists, who in turn do not have to trouble themselves with sociologists. This is all well and good if you are nestled safely, deep in the pithy heart of your field. If your problem is unquestionably about electrons or plants or human group behaviors, if you faithfully obey your Kuhnian paradigm, no harm can befall you.

What if you want to study an interesting problem that happens to lie in the white space in between the fruits? You might want to solve machine intelligence problems based on self-modifying human psychology or social interaction. Or you might be interested in how communities of science behave under forces of both collaboration and competition. These problems don’t fit neatly underneath the hardened, protective rinds of our departments, which protect us from intrusions by heathen areas of study. You will have to endure extra paperwork, or form your own laboratory for media, or resist the scorn and derision of well-established fields who compete for limited funding and publication space like corporations compete for capital and market share. It is nothing personal, weird interdisciplinary problems, we just have to look after our own. Especially in this recession.

But given the space of all possible problems, the amount of this interdisciplinary whitespace represents a significant long tail of worthwhile science. Inaccessible and fragmented, these “in-between” fields and the researchers interested in such pan-fruit collaboration might look to the example of e-commerce. In the same way Netflix and Amazon created new markets for distributed, niche tastes in products and entertainment, there are new hybrid fields like bioinformatics (computational biology), biophysics (physical biology?), and other fields formed by combining biology with some other existing field (biological weapons are the atomic weapons of the 21st century). But unlike rentals of obscure DVD cult classics and mountain-climbing books by Jon Krakauer, there is no easy way to automatically break down the barriers between academic departments. You cannot simply say “this problem is interesting” and judge it on its own merits. You must justify how relevant this problem is to this or that established field, or you must go it alone and start a new conference or publication.

Some of these interdisciplinary problems might be:

  • What is the best way to divert funds from a Swiss bank or Cayman Islands tax shelter in order to fund my research?
  • Should I conceal my lab in the crater of an active volcano, against the prying eyes of Google Maps?
  • Can ransom threat futures be securitized and traded on a public market?
  • How can I save on long-distance calls when delivering ultimatums to world leaders?

These problems cannot be conveniently labelled, except as mad. Kids interested in science have been defanged and sequestered such that tomorrow’s young Feynman and little von Neumann would never entertain a career in global domination. They are too busy justifying their Defense Department contracts to see the connection between currency manipulation and electronics manufacturing, and whether they can get in on that action. Or stop it. But if not mad scientists, then who? To be continued…

(As a tangent, note the resemblance of the citrus fruit slices above to Hamming balls covering the word space of error-correcting codes. The center of a ball represents a valid code word, and the error correcting code can uniquely decode errors which are within a certain radius of each code word. But what about errors which lie in the white space between Hamming balls? It turns out that by relaxing the constraint of a unique solution, we come up with list-decoding, which returns a list of probable code words for such an ambiguous error.)


in the style of a hideous man

•22 March 2010 • Leave a Comment

Shuttling up and down the metal, his fingers played on valves and keys, push-rods and joints welded in a maze.

“Because the air column, this part here, that determines the pitch. So to raise an octave, you have to stop it—here. That’s what this key is for. But the octave is slightly different for every note, and once you get to G there is a completely separate stop. It’s a little kludgy, yeah…”

Uh-huh. It was the Industrial Revolution, and some engineer type with an extra welder decides he is a fucking genius, he can make a new musical instrument. And it is the most complicated, bizarrely constructed—and I would use the word ‘Byzantine’ here, but I’m sure the Byzantines had much better-designed musical instruments and the comparison would not be flattering to them—as if they improvised with a malleable bronze sheet, a punch, and a file, themselves instruments in a trio of manufacture, as if they drew lots, or were steered by inauspicious signs in the flight of geese, or went at it by iterations. After they snaked the first layer of levers and buttons and their interconnecting lines of force, they would have to repair their inevitable deficiencies, the pushes and pulls would need adjustments, hence the bolting and screwing and duct-taping on of a second layer of dull, golden knobs, bumps, doodads, tchotchkes, hood ornaments. And this layer of repairs would also be so abstruse, so convoluted, that it would require another wave of error correction. It probably dispensed PEZ also, I never checked.

“And Blackman, he had this completely crazy new idea, an octave key for every note. So that it’d be exactly an octave see, you don’t have to fix it up by shaping your mouth differently or blah blah, embouchure, blah, spit-valve, blah, this one time in band camp.”

Well, that was the gist of it anyway. I remembered the words that affected me the most, and he was always on about that fucking embouchure. But he would say it like a French word, “am-boo-SHOOR” with a hacking noise at the end as he imagined they would say it at the court of Louis XIV, the Sun King’s marching band or Versaille jazz ensemble.

“Completely essential, though, if you start a modern jazz ensemble. Essential tone color, indispensable. Sine qua non.”

But who invents like this anymore? Who freezes any new features at 75%, standardizes it, and then goes to mass production? And when the standard model is not enough to satisfy the market for this labyrinthine aesthetic, to release different models to cover the entire, imperfectly-stopped octave range, the alto, the baritone, the bass, the soprano, who the hell knows how many there are.

“Once you know how to play one, you can pretty much play them all. I mean, you wouldn’t normally, you never see these stranger cousins outside of some serious brass geeks. Those guys are show-offs.”

At some point we draw a dividing line after which new instruments can only be old instruments hooked up to an amp, maybe with some wah-wah, a little distortion here, a little saturation there, or they are your MacBooks running Reason with a USB mixer. More knobs, more sliders, diamond tips on vinyl which recorded the originals, CDJs in digital imitation of same. Because we had discovered them all, there were no more to be had, like alien artifacts planted during pan-spermia, or constructed by our enslaved ancestors via stargate. Henceforth, it will only be samples, played on a Casio keyboard, and remixed. Endless copies of facsimiles of varying degrees of verisimilitude and THX remastered special editions, new characters, changed endings, half-hearted post-facto attempts to improve on something whose very charm was its imperfection.

“Yeah, I guess you could put an electric pickup in there, but why would you? What is this, copycatting stringed instruments? No way, I would never do it.”

Truly, a feud to last the ages, on par with Sox v. Yankees, Roe v. Wade: band kids versus orchestra kids. Strings stopped evolving long ago, when crude noise-makers were carved out of wood and stretched with cat-gut, and made to whine and screech with horse-hair rubbed with pine resin, and other disgusting plant/animal byproducts. Pfft. Barely more advanced than banging stones together. Consider the banjo, the viola, I rest my case. But those quaint mandolin and ukeleles have no innovation such as our little monster, our jeweled and plated beast.

In the future, they will keep the original in Paris, in a vacuum along with the platinum-iridium cylinder that weighed our cattle and metric monster trucks during the Middle Ages. Take your kids to see it. They will demand what evidence remains of Bill Clinton’s musical talent, they will want to know what mourned the passing of hard-boiled detective noir, they will ask what if H.R. Giger dreamed in brass. The last musical instrument made by the race of man, that’s what.


•14 March 2010 • Leave a Comment

Holding two contradictory ideas in your head is a useful trick for humans, and it can help explain some of our behavior as polypsychic creatures. That is, our minds and personalities are complicated enough to be considered a collection of different parts, competing and collaborating with one another to put up this pretense of a unified person to the world.

Lateralization of brain function and lobotomy case studies already show biological evidence that a single human being can already be considered two interacting brain halves. So it is not a far jump to consider that any sufficiently complex system, with many independent yet communicating parts, can become inconsistent. In fact, for any interesting complex system, it would be unlikely and strange if it remained completely consistent over time, especially since we know inputs from the outside world can be random and imperfectly communicated.

For example, I watch a lot of (Internet) television and read a lot of books, which is influenced by what my friends recommend. I believe that traveling the world is great way to gain a wider perspective on my life, learn about different cultures, and keep in touch with these friends. However, I also believe that modern air travel is unsustainable, outputting huge amounts of carbon into the atmosphere and contributing to climate change. So what do I do?

Well, I mostly fly anyway and then feel guilty about it, maybe purchase some bullshit carbon offsets, maybe catch a matinee of cap and trade kabuki theater (thanks to Caitlin for this link). Another way is to reconcile the contradiction on a more indirect level, like Al Gore, and consider some air travel necessary to fight global warming. The world would be a much poorer place without the exchange of people and ideas, and if I am flying to Haiti to do disaster relief work, the benefits of my travel might outweigh the downsides. And how are we supposed to reach international agreement on carbon limits without flying our national leaders to Copenhagen? How will we get grassroots support for emission reduction unless the former Vice President works out of a sprawling palatial home, wins a Nobel Peace Prize, and flies around evangelizing his books and Powerpoint presentation? We would be fools not to allow these exceptions to the rule.

Well, this reconciliation may work for Gore, but I don’t find it particularly satisfying. I could indirect one more step, pop one more frame off the stack, and redesign my life to slowly circumnavigate the globe using only pre-industrial means. If both travel and reducing carbon emissions were truly important to me, then there is a way to make them consistent with each other. Like most paradoxes, the contradiction disappears from a wider perspective. Like going back in time to kill your grandfather. It was will being have turned out that you are actually Bruce Willis in 12 Monkeys. (This is analogous to the Church of the Larger Hilbert Space, a view of quantum information that any “spooky” “waveform-collapsing” is actually non-spooky and non-collapsing in some larger state which contains you and the measurement). In some context, crossing the globe and saving polar bears is completely consistent.

It just happens to not be the context where I maintain a stable residence in Seattle and complete a Ph.D. program in computer science. I can’t have it all, not at once anyway. But by maintaining this contradiction in my mind, I will curb my excessive airplane hopping until such time as I can become a nomadic engineer-wanderer. By deferring a complete consistency check, I can accomplish more of my goals overall. Or I can decide to have fewer goals, and maybe travel isn’t that important after all. Although it is impossible for human beings to be completely consistent about anything, it is very human to learn by resolving or reframing contradictions.


•7 March 2010 • Leave a Comment

Recently, I heard Jared Diamond speak at Benaroya Hall (thanks to Luke, my favorite barista, who invited me). He is the famous author of Guns, Germs, and Steel, and unofficial spokesman for research tourism in Papua New Guinea. This book probably contains the most interesting ideas about the rise and present shape of human societies ever, but the writing is a little number heavy and dry. Somehow, this exact same style is way more entertaining in a live performance due to Diamond’s droll, kiwi-influenced accent. Now like every good author, he is out pimping his new book, Collapse, about the corresponding fall (or not) of societies.

The one word that was reinforced throughout the whole evening was geography. This word evokes in me vaguely sterile experiences of looking at colored maps in middle school, memorizing states and capitals, and writing reports on faraway places exclusively through reading the encyclopedia. But when Diamond uses the word geography, he actually means a more inclusive definition that encroaches on ecology, sociology, anthropology, ethnobiology, geology, history, economics—fields of knowledge which we tend to isolate into their own silos, the better to quantify them and practice a reductionist brand of science. The older and more colorful mixture of knowledge about the Earth’s regions, inseparable from their location on the planet, is famously popularized by the National Geographic Society.

We can all agree that National Geographic is sexy. After all, they brought us awesome photographs of terra cotta warriors in China and the Afghan girl and zebras mating, or whatever. Then why is geography is the least sexy subject in school? In fact, less than 1 percent of universities in the U.S. have a department of geography. Once you have reduced this field to cataloguing Anglicized place names, it hardly seems a worthwhile academic discipline.

As usual, pundits and futurists are to blame. It has been popular recently to make fun of location as irrelevant, a holdover from the last century when transportation was slow and we weren’t connected by a series of tubes. Those poor prehistoric sods, they couldn’t even outsource their call centers to India or import tasty, endangered sea bass from Chile. Thank god Thomas Friedman is around to tell us about globalization. For economic reasons, we no longer need to be bound by the accident of location. We can have everything, be everything, no matter where we live or were born. I can’t wait until we are all brains in a vat or upload our consciousnesses to the Internet after Brother Kurzweil’s rapture.

Unless, of course, oil stops being free and shipping containers don’t move themselves. Or the game of third-world labor arbitrage is balanced out by Asian nationalism and the need for authentic human relationships. Or some backhoe runs over the power cord to our cerebral aquarium or the fiberoptic line of our global socialputer. Stupid backhoes.

What about disregarding location for intellectual reasons? In Richard Dawkins’s 1991 Christmas lectures, he imagines the scenario where space aliens land on Earth and give us a report card on our progress as a species. In the arena of science and mathematics, they might pat us on the head patronizingly and say “Eh, you’re on the right track. Keep it up, you’ll get there some day.” After all, the wavelength of the hyperfine transition of hydrogen is the same everywhere in God’s creation. Why else would we have put it on the Pioneer plaque? But these aliens, they will either be uninterested or disappointed in our religion or culture as small-minded and provincial. After all, what does it matter to the universe how a tribe worships its deity on a small planet in a tiny solar system in one of countless galaxies?

As much as I heart Dawkins, we are stuck here on this spaceship earth for the immediate future. And as Jared Diamond teaches us, our cultural and religious beliefs are not just arbitrary and irrelevant local phenomena. They can be traced to the shape of continental landmasses, the latitude of our evolution, the direction of human migration—gross physical characteristics of our planet that are tectonic in scale. There are parts of reality that can be described by mathematical equations, and there are parts that can only be explained in the context of their time and space, their real-estate prices and political parties, their history and geography.

I like history too, by the way.

hole punch tests

•14 October 2007 • 1 Comment

hole punch test

julie: I think biological differences btw men and women are exaggerated.
me: But they *are* different. e.g. men have better spatial orientation.
julie: what?! that’s bullshit.
me: no, on standardized tests, they fold these sheets of paper up, punch holes in them, and then ask you what they look like unfolded.
julie: blah blah they look like a bunch of holes.
me: see, that’s the woman answer.
julie: I bet a man wrote that test.
me: Obviously. We’re not going to trust women with their crappy visualization to write our punch-hole tests.

More or less. I am ever watchful against the zeal of neofeminists to mow down mountains in order to make a level playing field. However, in this case, I should perhaps clarify what I think is useful about these tests. In the 1980s, among hair bands and regrettable pop music, there was a trend in standardized intelligence tests that emphasized punching holes in folded pieces of paper and rotating shapes as the sole arbiter of spatial cognition. Mostly by the ETS, the same scam artists who run the SAT/ACT/GRE pyramid scheme. I can only find one example online, on a dental admission test thanks to Google Books, but trust me, these used to be all the rage in bullshit academic circles.

Thankfully, someone has already done the hard work of writing a Ph.D. thesis about gender differences in spatial cognition using these tests. The newly doctored Ho Chun-Heng has described three popular tests in Section 3.1, including the much-vaunted hole punch test (which tests spatial visualization) and the card rotation test (which tests speeded rotation). I stand corrected. There is no significant difference between genders in spatial visualization, but among design majors, there is a gender gap for speeded rotation. This suggests that on average, in a population of men and women pre-selected for spatial interests and abilities, men will better be able to turn screws facing away from them, give directions from an upside-down map, and play Tetris.

Julie is correct that these tests are probably written by men and have an inherent bias for tasks for which men may be better adapted biologically. So the use of the word “intelligence” here is loaded and misleading. While I consider all of my statements above as true, they are not the complete truth. I have a continuing interest in identifying particular tests or mental tasks which show a gender difference in the opposite direction (i.e. women score higher than men). Gender differences in mental tasks should be not be invented where they do not exist. However, ignoring differences when evidence suggests otherwise would only maintain the status quo in educational methods.

heart of darkness

•7 September 2007 • Leave a Comment

While NPR was extolling the virtues of 4-and-a-half-star badass David Petraeus this morning, the real jewel comes to light. As a lieutenant colonel in 1991, he was accidentally shot in the chest by an M-16 during a live-fire exercise in Kentucky. They medivac’d him to Vanderbilt and scared up the best heart surgeon in Nashville to operate. Little did anyone suspect at the time that this medical marvel would later go on to become Bill Frist, the honorable senator from Tennessee and Republican majority leader.

Frist is still certified to perform general and heart surgery. Imagine, if you will, this disaster: Dick Cheney collapses at his podium during some diabolical Democrat filibuster/all-night slumber party. His aides are frantic. “Is there a doctor in the house?” Frist leaps to his feet, and in the confusion, slips into the nearest phone-booth or men’s restroom. Here he rips open his double-breasted Armani pinstripes and starched, white button-down Oxford, to reveal blue scrubs. Sliding a mask over his face to protect his secret identity as a mild-mannered Congressional powerhouse, he rushes back to the prone figure of the Halliburton CEO.

“Scalpel! and the heart of a wild boar, stat!” he roars. Conservatives thank their christian god and cheer him on, sissy liberals across the aisle faint at the sight of blood or swoon at his manly charisma. Hours of feverish work later, as C-SPAN cameras roll, Cheney squeals to consciousness. Frist collapses in exhaustion, he is fawned over by sexy nurses, someone hands him a stiff drink. The country is saved.

Oh right, Petraeus really is a badass. And overcompetitive, too, whouda thunk? I didn’t mean to divert the spotlight. But really, a heart surgeon. I mean, seriously.


•3 September 2007 • Leave a Comment

Q: What are your social blindspots?
A: When changing conversational lanes, I send signals but I don’t really know if the coast is clear.
Q: Have you tried reflecting on this problem?
A: Yeah, but it’s difficult because this issue is closer to me than it seems at first.
Q: How do other people feel about your sociopathic tendencies?
A: Good? Or maybe bad. I’m not really sure. Hey, fuck you, shithead.