I am pretty much required to blog about this piece in the New York Times entitled “They Tried to Outsmart Wall Street”: “they” being physicists who left science for Wall Street. And the not-so-subtle implication of the title is that we failed to outsmart Wall Street (and possibly wrecked the economy in the process). To that I say, in the words of Bart Simpson: it was like that when I got here.
More seriously, while I have no doubt that there were crappy quant models out there that contributed to the current crisis by maximizing short-term gain over long-term risk, this was true all the way up and down the chain and the quants don’t deserve any more of the blame than anyone else. Quants respond to incentives the same way everyone else does, and the compensation structure on Wall Street can incentivize immediate profit and deferred risk. (My employer is trying to curb this effect by instituting a clawback provision on bonuses; I don’t know how widespread this is, but it seems like a good idea to me.)
The Times article is curiously focused on ex-physicists, as if quants don’t come from any other fields. In my ten months on the Street I’ve met quants from a broad range of science and engineering fields, and physicists aren’t a majority. That might be a peculiarity of my department’s hiring practices, with physicists being much more common elsewhere, but I’d be surprised. Anyway, Kevin Drum noticed this too and wonders why physicists are so suited to quant roles. He has a theory that it’s the culture:
Even among the number crunching set, physics has a reputation as the most aggressive, male dominated branch of geekdom: only 14% of physics PhDs are women, the lowest of any of the sciences. (Math is pretty male dominated too, but pales compared to physics: 29% of math PhDs are women.) If the first thing that “aggressive and male dominated” reminds you of is the big swinging dick world of high finance, give yourself a gold star. Call this the testosterone theory: physicists are attracted to Wall Street because they like the atmosphere.
I don’t think this is right: the atmosphere in a typical physics department is nothing like the stereotypical Liar’s Poker trading floor that Drum is alluding to. To the extent that the environment I work in is like academia, it’s because I’m lucky enough to work with a group run by ex-academics rather than people with a typical trader’s background. Instead, what Drum calls the “affinity theory” really is the right one. The work I do now is a lot like the problems I worked on as a physicist. It’s not just (as Drum suggests) about math; it’s about the ability to work with huge data sets and make sense of them, and to find signals in a noisy system. This is a much bigger factor than testosterone levels.
I filed my dissertation this morning; I am now Dr. Arcane Gazebo. (Well, technically the degree isn’t conferred until Thursday when the semester ends, but whatever.)
The main result of the entire thesis comes down to a single plot, shown below. This isn’t the “explain my thesis” post so I’ll just say that the plot shows our ability to control the coupling energy between two qubits by applying a bias current to our readout device, hence the thesis title Solid-State Qubits with Current-Controlled Coupling. The solid curves are calculations based on device parameters and the dashed curves are one-parameter fits.
Now these points of data make a beautiful line…
If anyone needs me this evening, I’ll be at Triple Rock.
I thought our lab was a mess, but it could be worse… via Chad Orzel, here’s a chemistry professor (at UT San Antonio) whose lab had to be forcibly cleaned by the university:
“Clean your room or get out!” Words from a frustrated parent to a messy teenager? Not quite. The mess-maker in this case was a chemistry professor at the University of Texas, who ignored repeated warnings to clean up his dangerously cluttered lab space. When University officials decided to clean it themselves, the professor caused such a disturbance that campus police had to lead him away in handcuffs. The professor was eventually fired, which prompted a lawsuit claiming that the University retaliated against him and denied him equal protection.
The legal opinion notes that apart from the problems in the lab, the professor’s office was an “extreme fire hazard”, which still puts him a step below the physics professor here at Berkeley who actually set his office on fire. In any case, this makes me feel better about the disordered state of our lab. We cleaned it only a few months ago but it returns rather rapidly to equilibrium.
(I also want to point out that the legal blogger linked above is evidently a fan of Arrested Development, and has chosen the obvious pseudonym to use on his law blog…)
Via Shellock, there’s a fascinating post at Gene Expression on various findings that show that intelligence is correlated with delayed sexual activity. There’s a lot of interesting stuff in the post and I encourage reading the whole thing, but I want to point out the results I found surprising. Not because they go against stereotype—they actually confirm “science nerd” stereotypes, but I had convinced myself that these were just stereotypes without much basis in fact. These numbers indicate otherwise: (emphasis in original)
By the age of 19, 80% of US males and 75% of women have lost their virginity, and 87% of college students have had sex. But this number appears to be much lower at elite (i.e. more intelligent) colleges. According to the article, only 56% of Princeton undergraduates have had intercourse. At Harvard 59% of the undergraduates are non-virgins, and at MIT, only a slight majority, 51%, have had intercourse. Further, only 65% of MIT graduate students have had sex.
I was quite shocked that the numbers were this low; I obviously know a lot of grad students, and though I haven’t polled them on this subject, I would have guessed a much higher percentage. (I’m not chauvinistic enough to suggest that MIT grad students are less sociable than those at Berkeley—I expect the populations are pretty comparable, at least in departments like physics.)
However, I may be thinking too narrowly in terms of the stereotype of scientists who are virgins because they are socially maladjusted. (There are people like this in the community, but it’s a small fraction.) The Gene Expression post lists a number of other possible reasons this could appear as an aggregate effect, and argues for a few of them as contributing factors. (At an individual level, of course, it will be strongly path-dependent.)
One factor that wasn’t mentioned there is culture. This could manifest in at least two ways. The first is that a substantial fraction of grad students in technical fields are immigrants from cultures that are much more sexually conservative. Thus, even if these students themselves don’t hold conservative views, they may be less likely to have had sex. The second is that the culture in academia seems to me to be less sexually charged than in other spheres. This is not to say that it’s sexually restrictive—as the Gene Expression post points out, most academics hold liberal views about sex—but it’s less focused on going out and getting laid than, say, the Late Night Shots crowd. Our lab’s monthly board game nights aren’t terribly conducive to hook-ups (although surprisingly conducive to drunkenness).
Anyway, this might explain the results of the academic polls, but the original post is concerned with correlations with IQ rather than academic achievement. A logical extension would be to look at people in other intellectually-demanding disciplines, like law or medicine. Would the numbers be similar? My guess is no, but I may be stereotyping again.
As I start to see the light at the end of the grad school tunnel, I’ve been contemplating more and more my various options after I finish. The most obvious one is to go on to an academic postdoc, with the aim of eventually getting a tenure-track professorship somewhere. (Other alternatives are various industries or finance.) At the moment I’m leaning strongly against an academic career, which has lately seemed unappealing for a variety of reasons.
A major such reason is the fact that there are many more applicants for tenure-track positions than there are positions available, so that after slaving away for several years as a postdoc (generally considered to be an awful job) I’d be lucky to be offered a position anywhere. It’s a job market that’s extremely unfavorable to applicants, and having seen the stress and unhappiness it produces in the postdocs I’ve met, I am thinking I should look at other options.
One corollary to the scarcity of academic jobs is that I would have to take whatever I can get, meaning that I will have basically no choice over where I live—the institution that offers me a job could be anywhere in the country, urban or rural, coast or inland. And I’ve realized that where I live really is important to me. I like living near enough to a major city that I can take advantage of the cultural and economic diversity. Furthermore, I want to live in a walkable neighborhood where essential goods and services are close by—not just for conservation reasons, although this is certainly part of it, but because I’ve found firsthand that it brings a definite improvement in quality of life. (This, of course, is also only possible in or near a major city, and only in certain cities that are planned this way.)
And on an emotional level, I’ve found that I don’t want to leave the Bay Area. This surprised me, because (possibly due to my migratory upbringing), I generally feel like I need to move on every few years and explore a new place. I’ve tried to ascertain why I might have a special attachment to my current location: certainly I don’t want to leave my friends, and I like my current neighborhood, but I feel like there’s something more than that. There’s a sense I have of being settled here, that where I’m living now is woven into the fabric of my life. I haven’t felt that way about anywhere else, but I’ve lived in Berkeley longer than I have any other single place (for a continuous span).
I’m not convinced that this is a good reason to want to stay here—I know that living in different places is an enriching experience for me, and there’s some attraction to going and exploring someplace new. But it will probably influence my thinking on career options.
I took my qualifying exam this morning. If you’d like to know it went, click here.
I thought about posting last night but this was pre-empted by the fact that the slides for my talk were unfinished (and also the Clarke group dinner). First I want to register a complaint:
This is how physicists (or maybe everybody) fill seating at conferences. The first people to arrive take the seats on the outside of the rows, and then fill in to the middle. This is really annoying when arriving in the middle of the session and having to climb over a bunch of people to get into the one empty seat. I am aware that this is a really lame complaint, but please, fill from the middle!
Now that I’ve got that out of my system: the last couple days were a blur of superconducting qubit talks. There’s a lot going on in this field, and most groups had three or four (10-minute) talks in a row to have enough time to explain all their results. One experiment I thought was very neat was this one from Terry Orlando’s group at MIT. In flux qubits like the ones we study, one can measure the temperature by sweeping the flux bias across the degeneracy point and measuring the population of the qubit states. Higher temperatures will give wider curves, as energies further away from the degeneracy point are more likely to be populated by thermal activation. When we measure this on our qubits we usually get something like 150 mK, mysteriously somewhat higher than the fridge temperature (roughly 50 mK).
What the Orlando group did was to apply an analog of laser cooling (as in atomic physics) to their qubit, using a microwave pulse to induce transitions that ultimately cool the system. As a result they were able to see these temperatures (as measured from the widfh of the qubit step) reduced by a factor of 100, from 300 mK to 3 mK. It was pretty impressive; I’m not sure how important it is for quantum computing or whether it’s something we should be doing with our qubits, but it’s a nice application of techniques from another field.
This morning I gave my talk, which was helpfully introduced by Frank Wilhelm’s talk immediately prior, in which he said something like “the really important development for scalability is what Travis Hime will talk about next”. So the pressure was on, but I think I did ok. After this was… more qubit talks, but I was mostly decompressing after finishing mine and didn’t pay as much attention as usual.
Tomorrow I go to see talks by other Clarke group members, including John himself. And then, an evening flight back to Berkeley.
Actually I spent much of today working on my talk instead of going to sessions. The superconducting qubit sessions start tomorrow morning and basically run continuously until Thursday evening. I did go to some talks in the afternoon, though, mostly in D2: Ion Traps for Scalable Quantum Computation. (In some sense this is our competition.)
Ike Chuang, who is a big name in this field, gave the first talk, which laid out the challenges in making a practical quantum computer with ion traps. Most of this dealt with error correction; according to Shannon’s theorem (or maybe a quantum information version thereof) it should be possible to build an error-free quantum computer out of qubits that do make occasional errors, as long as the failure rate is below some threshold. Unfortunately in some cases they’ve looked at this requires a prohibitively large number of operations, as many as 1020. One can try to implement various error-correcting codes, such as Shor’s or Steane’s, but certain operations that are needed for a universal quantum computer don’t work within these codes. And in fact Chuang et al. have shown that there is no stabilizer code that allows a universal set of operations to be performed within the code—one has to decode first before performing at least one of the operations.
The other talks in the session were less abstract, and thus harder to understand (since I’m not terribly familiar with this architecture). The talk by Slusher described a proposal for a VLSI-based scalable ion-trap based quantum computer, which seemed impressive, except I’m pretty sure this is the one Chuang mentioned that would require 440 watts of laser power to operate.
I skipped out on the last talk to go to D8: Superconductivity: STM of Cuprates and see what the group I worked in as an undergrad was up to. However, I haven’t thought about STM of cuprates for a while now and only had the faintest idea what they were talking about.
A tempting alternative for the end of the day was Session D33: Focus Session: Quantum Foundations II. It starts out as a perfectly normal session, but somewhere around 4:30 becomes the dumping ground for crackpots. For example:
D33.00014 : Do Particles have Barcodes?
If an elementary particle shown in Fig 2 of gr-qc/0507130 has an UNSTABLE quantum connection to the rest of the universe calibrated by nature in terms of Planck times, as also proposed in my separate MAR07 abstract, there exists a possibility that each particle has a barcode of its own. Instability implies varying periods of connections and disconnections of particles to the universe, which would be equivalent to the varying widths of white and black strips of commercial barcodes. Considering the high order of magnitude of Planck times in a second, each particle and the universe generated by its radiations may have their unique birth times registered in their barcodes. My quest for the cause of consciousness, in MAR06 abstracts, as an additional implication of physics/0210040, leads to the inquiry if these unique parallel universes are like the ones that give rise to consciousness as proposed by some physicists. With all due respect, the attempts to explain TOE of inert matter may not be attempts to explain one step to climb up on a stairway at a time. They may be attempts to explain only half a step at a time to on a stairway made with only integer number of steps. The search for TOE assumes such a theory exists. Mathematics has no barrels to fire bullets that can shoot down a non-existent bird. A Hamiltonian knows no consciousness, a missing ingredient of biology made of particles or vice versa, and of realistic TOE.
The talk after that one describes a theory of Atonic Physics [sic], which sounds like an outtake from Monty Python’s bookstore sketch.
Today’s amusing search request: should I make an outline slide for my APS march meeting talk?
My physics category archive is the second hit for this search in Google. This is a surprising query to see from (presumably) a physicist: an overspecific question phrased in standard English is not the most well-formed Google search. (Some search engines are designed to take queries in this form, but Google is not one of them.) Nevertheless, the searcher lucked out: the fifth hit is a set of slides on giving good scientific talks.
I’ll answer the question anyway in case anyone else is wondering. If it’s an invited talk, the answer is almost certainly yes—a 30-minute talk will cover enough different points that an outline at the beginning will help the audience follow the transitions. If it’s a contributed talk, with only ten minutes of material it may not be necessary. If the talk divides nicely into multiple distinct sections, it’s a good idea, but if it’s centered on a single result you probably don’t need it.
There’s a great post at Cosmic Variance about the cult of genius in physics:
During high school or college, many aspiring physicists latch onto Feynman or Einstein or Hawking as representing all they hope to become. The problem is, the vast majority of us are just not that smart. Oh sure, we’re plenty clever, and are whizzes at figuring out the tip when the check comes due, but we’re not Feynman-Einstein-Hawking smart. We go through a phase where we hope that we are, and then reality sets in, and we either (1) deal, (2) spend the rest of our career trying to hide the fact that we’re not, or (3) drop out. It’s always bugged the crap out of me that physicists’ worship of genius conveys the simultaneous message that if you’re not F-E-H smart, then what good are you?
I remember clearly the moment I found that physics was much harder than I realized (although I had no delusions of being F-E-H smart by that point anyway): it was Ph 106a. I was used to being able to pick up concepts fairly quickly, but the subtleties of advanced classical mechanics (and Goldstein’s textbook) eluded me, and it was a serious blow to my confidence that I really didn’t get it. I worried that this was a sign that all the high-level physics concepts would be beyond my reach. Obviously that turned out not to be the case; I just needed to work a lot harder to understand these concepts. It’s striking to me how rapidly the difficulty seemed to ramp up, but this may have been due to the way Caltech structured the physics curriculum rather than an inherent property of the subject.
Chad Orzel has a related point:
Too many people approach physics as if there’s some sort of Great Chain of Being, with the most abstract theoretical particle physics at the very top and low-energy experimentalists down at the bottom, just above biologists and rude beasts incapable of speech.
This drives me right up the wall.
There’s no inherent moral worth to working on more “fundamental” and mathematical physics. A lack of familiarity with algebraic topology is not a defect in character, or a sign of gross stupidity. Low-energy physics is different than high-energy theory, but not inferior to it.
This is something I noticed a lot as an undergrad—in my freshman class almost everyone who wanted to do physics was interested in high-energy theory; I was rare in actually being inclined towards experiment at that point. Part of it is that there’s a certain glamour to working on the Theory of Everything, and there’s an apparent elegance to a simple but widely applicable theory that makes the experimental world look messy and ugly by comparison. (Although in fact the Standard Model isn’t really what I’d call simple or elegant.) Furthermore, at roughly the freshman undergrad level the major contact with experimental physics is through high school or freshman physics labs, which tend to be pretty lame.
(So how did I end up wanting to do experiment at that stage? At the end of my senior year in high school I had the opportunity to do some labs on more advanced topics, and they were less structured than what I was used to—instead of the procedure being laid out explicitly, I was given a set of equipment and had to figure out how to use it to measure a certain parameter or figure out how something worked. Although it was still pretty far removed from the actual practice of experimental physics, it gave me a better sense of the kind of problem-solving involved, which I found I really enjoyed. Plus I noticed I was better at it than I was at theory.)