UW Daily Update Monday Nov. 23: Artificial Life and Effective Altruism

1.     UW Daily Update Monday Nov. 23

a.     Grab a egg sandwich at Motosurf, then take final physics midterm. It's fairly easy.

b.     Go to the SPS lounge and write my portfolio statement for Honors 100.

c.      The lunchbox seminar is about some small steps towards “artificial life”. Our professor has designed particles that bind together using the “lock-and-key” model. Larger spherical particles with semispherical dimples fit together very closely with smaller spherical particles. (These particles were large macromolecules made of many small linked monomers.) When a large particle and a small particle fit together, they bind due to what is known as “osmotic pressure”. The particles are floating in solution, so on their own they experience constant pressure from all sides, due to both water and larger molecules in solution. But when their surfaces get close enough that there are some water molecules between them, but the larger molecules can’t fit in the gap, the surfaces facing outwards experience pressure from both the water and the other particles in solution, whereas the surfaces facing inwards only experience pressure from the water, so the lock-and-key particles experience a net force towards each other and bind the rest of the way. This is called “depletion binding” since the inner large molecules need to be depleted before it can occur.

d.     But, I hear you ask, as I asked, don’t the larger molecules displace the water, so that when the larger molecules are forced out of the gap between the surface and they are replaced by water, the water pressure inside increases by the same amount the pressure due to the larger molecules decreases? The professor, being a physicist, not a chemist or biologist, gave a entropic explanation rather than a mechanical one. The total entropy of the system is lower if the larger particles are stuck together, because then there is more volume available to all the molecules floating in solution outside the particles, and this means more possible states for these molecules and hence more entropy. (The professor stated (I can’t give the derivation) that entropy = c*number of molecules*ln(volume available to each molecule).)

e.     Anyway, the professor used the larger particles’ ability to side-link to create chains of particles to seed the “reproductive” process by which the particles linked with the smaller spherical particles, which then linked with larger particles that naturally lined up in the same order; if the larger particles were small-dimple-big-dimple-small-dimple, they would link to small-sphere-large-sphere-small-sphere, which would in turn link to small-dimple-big-dimple-small-dimple. (There were multiple sets of locks and keys; a given type of smaller particle only fit one size of larger particle) These secondary chains would then split off, since the interactions between the locks and keys were relatively weak, and you’re left with the original chain and a copy. Sometimes a wrong molecule would bond, creating a mutated copy that would compete with the original one.

f.      The most remarkable thing that occurred during the seminar: a young woman studying biology, Rachel, whom I knew from Go club had come into the physics lounge to study with no idea that the lunchbox seminar was going on. I told her the topic of the seminar, and she was very excited. During the seminar, she jumped up to discuss a new technology, helical protein bundles, with the professor, since it would solve many of his technical issues. At first, he was skeptical, but she pressed on. I caught myself holding my breath at one point. After many questions, the professor asked Rachel for citations, which she readily agreed to provide. It was awesome.

g.     I discussed a little bit with Gorm, the tutor graduate student I had met at my first seminar, then left for lunch.

h.     I read a fascinating Ribbonfarm article at lunch, then went to CS class, where Reges gave the clearest exposition on abstract classes I had yet received. Reges is such a good lecturer that he strengthens topics I thought I had solidly covered.

i.       I dropped by the Commuter Commons, a place for off-campus students to hang, to try to get advice on how to get to the airport for my family’s Thanksgiving visit to extended family in California. I gained some good information and readied myself to take public transit the next day.

j.       Went to SPS lounge again, this time for Go club. Played a quick game of 9x9 with Rachel. I played pretty well, but messed up a corner invasion. Learned a little bit more about her—she hopes to go to grad school at Stanford, but fears her application will be denied since she can’t put her current research on yet. I want to tell Eleanor I’ve found a second Natasha J.

k.     Earlier in the day, I saw a flyer in the SPS lounge for “Effective Altruism Informational Meeting: altruism supported by evidence-based reasoning” in the Haggett South study room, run by one of my floormates Issa Rice. I decide to go, feeling thoroughly Leroy. I fear that the group will be dominated by passionate social justice advocates or people who advocate some stilted version of rationality to an irrational extent. But I get extremely lucky, and find a group that consists of very reasonable people. The head of the Seattle group, John, who came to help Issa with the meeting, is an Amazon programmer; half of the others are CS majors and programmers, while the rest are STEM majors. The regulars dismiss ideas that they think are irrational (in each case, I agreed) with politeness, but an almost politically incorrect refusal to hedge their statements with uncertainty they don’t feel. It feels very refreshing. Everyone has agreed on values to an extent that allows efficient discourse without side-stepping around direct assessments. We talk about the Effective Altruism group in Seattle, some interesting meta-charities, and future meeting topics. Afterwards, I talk with a mathematician, Tim, who tells me about a flaw with academia I hadn’t heard articulated: full professors are forced to simultaneously do research, teach, and manage an army of graduate students, and many people simply don’t have talents in all three areas.

l.       John provided pizza, so I go straight to my room without passing “dinner in the HUB”. I finish up my portfolio website and statement for Honors 100 and turn it in, get Rick’s, do laundry, print my boarding pass, and write this log.

UW Daily Update Saturday Nov. 22

1.     UW Daily Update Saturday Nov. 22

a.     Sleep in.

b.     Work on CS assign in lounge, then head to Which Wich and grab a sandwich.

c.      Use one of the bike storage lockers as a standing desk and get CS assign working outside.

d.     Play some badminton.

e.     Meet Dad and Mary, go to EJ’s Burgers on Ave.

f.      Go to football game. Spend the first half in the student section with Jamie and other floormates (Donovan, Katie, Alex). Spend the second half with Dad and Mary. We destroy OSU.

g.     Say goodbye to Dad and Mary by dorm. Refine CS assign. Eat some peanut butter and gorp and study for physics midterm.

h.     Write this log and hit the hay.

UW Daily Update Wednesday Nov.19: Lab tours!

1.     UW Daily Update Wednesday Nov. 19

a.     Wake up, eat peanut butter banana, go to physics. Talk about non-inertial frames of reference (if I measure an object’s motion from an accelerating frame of reference, it looks like it’s accelerating in the opposite direction as my frame even if there are no forces on it. Thus it doesn’t remain at constant velocity if there are no forces on it, and it violates the law of inertia.) Interesting stuff.

b.     Do CS hw.

c.      Do last physics lab on more angular quantities.

d.     Meet up at the SPS for lab tours! Yay! We trek over to the previously unidentified building below Haggett that turns out to be a rat’s nest of physics labs.

e.     After waiting 20min for the tardy professor, a grad student shows up and takes us to the Axion detector experiment. The Axion is a hypothetical particle with specific properties that might make up dark matter. We can’t see it unless it decays, and it decays very slowly. But we can force it to emit an electron when it hits a strong magnetic field. The experimenters have set up such a field and a detector. However, the axion will only emit an electron when hit with a field of the right frequency, which depends on the axion’s unknown mass. So the researchers have to comb through all the frequencies corresponding to the range of likely masses looking for tiny perturbations in the magnetic field owing to the emission of an electron. To distinguish these perturbations from thermal noise, the experiment has to be cooled to a few Kelvin (a few degrees above absolute zero) with liquid helium.

f.      Despite the sophistication of the experiment, the workshop looks completely hacked together. The grad students talk about doing the plumbing themselves. Stuff just sits on tables where it could be broken. Racks of electronics surround the command center: a folding table with two computers and a shabby folding chair. It’s pretty neat.

g.     Afterwards, we head back to the physics buildings and explore some of the labs there. All the labs are relatively small. Every surface and shelf is crammed with random-looking goo, whether it’s optical equipment and screwdrivers, samples of glass-like materials, bottles of chemicals, or giant specialized boxes of hardware. In one lab, a liquid helium cycling pump keeps up a constant whir-whir-whir that must drive experimenters crazy. The research is neat—magnetometers used to image biological structures, electronics made of thin sheets of graphene invisible to the naked eye.

h.     We head to the physics building and hear a panel of professors and grad students talk about how to get involved in research. It sounds like interest and technical skills trump big-picture knowledge of the science—good for me, who may not be majoring in physics.

i.       I hustle back to my room for a quick Skype with Grace, then head down to the 8 for dinner, then back up to the room for an evening of homework.

I write this log and go to bed. 

UW Daily Update Monday Nov. 17: Low-dimensional materials, x = change(x), and Go challenge

1.     UW Daily Update Monday Nov. 17

a.     Armed with peanut butter and raisins stuffed in my coat pockets, I head to the Motosurf truck and grab a very reasonably priced, generous serving of oatmeal. I head to the physics tables and chow down.

b.     Class feels like mostly review.

c.      Work on lecture homework. Get bogged down in loads of erroneous algebra. The problem is simple in concept and could conceivably be done any number of ways, but the art of the thing is to frame the problem in a way such that symmetries cause forces to cancel, resulting in simple math. I do it the hard way, and even then I could have solved it, but I make annoying mistakes.

d.     Walking to the Lunchbox Seminar, I mull over the advantages my awesome code editor has given me and resolve to procure a copy of Mathematica, to see if an interactive environment will do as much for math as it does for coding.

e.     I go to the engineering website, and Mathematica is free, but I have to wait a couple of days for student verification information to be processed, since I restricted release of my student directory information. Frustrating!

f.      The Lunchbox Seminar is fascinating as always. It’s the first time we have a female lecturer. She’s working on the properties of so-called “low-dimensional” materials like graphene, thin sheets of carbon 1 atom thick. As always, I feel like I need to have a better grounding in quantum mechanics to understand what’s really going on, but I get a lot of interesting stuff.

g.     Her research includes the following:

                                                 i.     Putting molecules on sheets of graphene to understand how the electrons react.

                                               ii.     Using a capacitor to add excess electrons to thin sheets of material, then observing how properties like the distribution of charge on the material change with instruments that take advantage of quantum tunneling of electrons to measure average charge.

                                              iii.     Sandwiching together sheets of different materials and figuring out how these very precisely defined contact interactions affect the materials.

                                              iv.     Measuring the conductivity of sheets of graphene, which usually contain defects which result in (resistance? Without the defects, would graphene be a superconductor with negligible resistance?)

h.     Graphene can be synthesized in neat ways.

                                                 i.     You may have heard this already, but pieces of graphene can be obtained by applying adhesive tape with just the right amount of stick to graphite and peeling off the tape. Then when you stick the tape down on a substrate more attracted to the graphene than the tape is, you can deposit the graphene on the substrate. Graphene obtained this way actually exists in square-micron size blotches of pure graphene interspersed between pieces of carbon several atoms thick, and places where no carbon was deposited. But the blotches of graphene can be detected and isolated for study.

                                               ii.     You can react a sheet of copper with methane (CH4), then use hydrofluoric acid (I think? Don’t really understand this) to react away the hydrogens, leaving just the carbon on the copper in very large sheets of pure graphene. Sheets the size of tablecloths have been synthesized this way.

                                              iii.     Graphene can be cut atom by atom using advanced equipment. Some catalysts actually cut graphene naturally along regular patterns. (In the presence of nickel ion, graphene severs jaggedly J )

i.       Neat stuff! Also, if graphene between a charge source and a charge drain is cut so that the middle is slimmer than the ends, like a constricted portion of a river, the electrons will behave as they would in a river, speeding up in the constricted portion (or so I gathered).

j.       After the Lunchbox Seminar, I eat lunch at the HUB (two small slices of pizza is not enough for me). Wander into Hannah and chatted with her while I ate.

k.     Go to CS, where we discuss parallel processing as a side topic and some list methods I’d already seen.

l.       Return to dorm and get computer science assignment working. Resolve some inner confusion about whether I need to use a pattern that’s a fundamental feature of C++, but a hack in Java. If I pass a variable pointing to a list as a parameter to a method in Java, and tell that variable inside the method to point to a different list, the actual variable passed in (outside the method call) doesn’t now point to the different list. To get this to happen, I have the method return the different list, and tell the variable pointing to the original list to point to the result of the method where I pass in that variable. This is known as the x = change(x) pattern. I figure out where I needed to use it and where it was extraneous.

m.   Go to Go club. Play a training game and learned some neat stuff. Rachel challenges me to a game for next meeting, saying (intimidatingly) she enjoys judging people’s strength.

n.     Eat pizza and read physics at the 8.

o.     Play badminton. I feel like I’m really improving!

p.     Return to dorm, shower, do physics hw, write log.