Recent Interesting Science Articles (September 2013)

It’s time for our monthly round up of interesting science stuff. It’s a pretty thin month unfortunately:

  • Let’s start with a fairly arcane article. It’s from Quanta Magazine and is about the discovery of a geometric construct that is called the amplituhedron. According to physicists, its discovery vastly simplifies the calculation of particle interactions. Along the way however, it does away with locality, which says that particles interact at specific points in space-time, and unitarity, which says that the sum of all quantum mechanical probabilities add up to 1. These are fundamental concepts in current quantum field theory and this discovery not only undermines these assumptions, but challenges the notion that space and time are fundamental constituents of nature. It suggests that space and time are emergent properties instead.
  • Next is a link to the actual research paper, which makes for rather heavy reading. It appears in PLOS ONE and is about orangutans may have better abilities to plan for the future than anyone ever suspected. It seems that when male orangutans travel long distances, they emit long calls which are used to communicate with other orangutans, and specifically may attract female orangutans and repel male rivals. The researchers found that the direction of the long calls emitted at dusk corresponded well with the direction that the orangutan chooses to travel in on the next day, suggesting that the animal is mentally capable of planning out its trip well beforehand and acting out on its plans.
  • Finally here’s a link that’s been going out to everyone interested in science. It appears in the Smithsonian and is about the discovery of the first and so far only functional gearing system found in nature. The gears were found in a species of long hopping insects. The gears lock their back legs together, allowing the two legs to swing at the precise same moment so that the insect jumps forward. It seems that this wasn’t discovered earlier because the system exists only in juveniles of the species as adults fail to regrow the gears as their skin molts away. The speculation is that the gears are too fine and fragile a structure to repeatedly regrow and even a single broken gear tooth would render the system useless.

From the Big Bang to Dark Energy

The latest Coursera course I’ve completed is From the Big Bang to Dark Energy by Hitoshi Murayama of the University of Tokyo. It’s a physics course and the first non-computer science one I’ve taken since Introduction to Finance back when Coursera first started up. The main reason I took it, apart from the obvious one of firming up what was only a very loose understanding of some of the most important concepts in modern physics and cosmology, is that at four weeks long it’s fairly short. I am due to travel to Greece for nearly two weeks in October and I don’t want to make any serious commitments time-wise until that is over.

As its title implies, the course is a broad sweep of the history of the universe, covering how the universe began and likely scenarios on how the universe will end. This is very much a popular science course tailor made for Coursera rather than a real University of Tokyo physics course so the professor delights in framing the issues in terms of human interest. How and when was the Earth formed? What are humans and all other form of life on Earth made of? How does the Higgs boson relate to everyday life? And so forth.

Professor Murayama is remarkably eloquent and despite having a noticeable Japanese accent when speaking English, his articulation is perfect. Combined with his willingness to insert interesting anecdote at opportune moments, this makes the lectures engaging and relatively easy to understand. As usual, the lecture slides are provided as a separate download in PDF format (the file sizes are quite large!) and the videos themselves look professionally produced, so there’s no faulting the quality of the course materials.

The original course information page lists some calculus and high school physics as prerequisites. In reality the course seemed to operate on two distinct levels. There are two quizzes every week with one week being easier and focusing on pure theoretical understanding, while the second quiz is somewhat and required some mathematical calculation to complete. The harder quizzes look intimidating at first glance but are actually very approachable once you take the time to understand what they are trying to teach. It appears however from discussions on the forums that even this was too much and many students couldn’t complete the harder quizzes, prompting the course organizers to comment that future offerings of this course will likely make a explicit break between an easier, more mainstream version, and a harder one requiring more mathematical knowledge.

For my part, while I did eventually complete everything, I couldn’t have done it without checking the forums. One reason is that my mathematics were rusty. But the main thing is that the quizzes ask to think like a physicist and that’s something that I’ve never developed. Rather than asking you to memorize formulas and apply them, the questions frequently wanted you to figure out the relationships between the variables given from their units and shift the variables around until you get the units and therefore the answer you want. Apparently this is called dimensional analysis.

One small problem that I found annoying is that the quizzes explicitly asked you to search for essential pieces of information yourself on the Internet, for example, what is the mass of the Earth? What is the mass of Mount Everest? What is the energy of the w boson? Given that different sources may give slightly different values, this made the search aggravating. They should just have bitten the bullet and given the values themselves. Eventually they did add a glossary and included a ton of extra information but it would have been nice for the course list such useful information early on. People like me who haven’t studied physics in more than a decade need some reminding that a Joule is the same as 1 kg·m2/s2 and a Watt is a Joule / s. There were also some tricky problems with rounding the answers in just the right way and presenting the answers in the correct format with scientific format so that the grading script would parse them correctly.

Still the intent of the quizzes is clear and I really liked how they tried to encourage students to think more about the subject matter being taught. As I understand it, the quizzes were entirely done by the course’s TA, Brian Henning. He even made a series of videos himself to help students with the harder quizzes and explain the mindset behind them. Unfortunately while his enthusiasm is certainly laudable, I didn’t find those videos particularly useful.

I did learn a great deal from this course. It’s great to finally have some concrete understanding of what is really meant by terms like Higgs bosons and dark energy etc. I was also interested in the details of how the physicists actually do their experiments and the thought processes behind them. One subtext throughout the course is a plea for public support for big, and necessarily expensive, science projects. It may seem somewhat self-interested of the course organizers, but it’s a cause that I support so I’m all for it.

One thing that does worry me is how little concepts like dark matter and dark energy are really understood. Despite all the time the course spent on these topics, they really come down to saying that because of so-and-so gravitational effects and other observations, we can infer that something is there. But they’re called dark precisely because nothing else is known about them beyond this. For this reason, it’s seems odd to me to teach them as accepted science. It seems to me, it could be equally likely that someone would come up with a new explanation for these observed phenomena that wouldn’t necessitate the presence of huge, invisible stuff in space. But overall, it’s a great course that I highly recommend to anyone interested in the topic.

A different prisoner’s dilemma

Here’s a fun riddle posed by a student in the recent Algorithms: Design and Analysis course from Stanford University I participated in on the Coursera platform. I’m recording it here for posterity so I won’t forget. It’s fun to collect such riddles. That might make for cool party tricks. Note that I no longer have access to the original post so all this is paraphrasing from memory.

Riddle:

There is a sadistic warden in charge of a prison with 100 prisoners. He wants to execute these prisoners, but after much persuasion on your part, condescends to play a game with them that could result in some of them from spared. Each of the prisoners is numbered from 1 to 100. A room is set up with 100 drawers, all also numbered 1 to 100. Inside each drawer is a slip of paper, again numbered 1 to 100. However, the placement of each slip of paper in the drawers is randomized.

The game is that each prisoner is allowed to visit this room, one at a time. He is allowed to open the drawers one at a time and may open no more than 50 drawers. If he finds his own number on a slip of paper after having opened 50 or fewer drawers, he is saved. He must return each slip of paper into the drawer he found it in and is not allowed to exchange the positions of the slips of paper.

As an additional concession, you are allowed to visit the room before any of the prisoners may do so and examine the contents of all of the drawers. Furthermore, you may perform one and only one exchange before the game starts. That is after having viewed the contents of all 100 drawers, you may choose any two drawers and exchange the slips of paper inside them. You may not make any other changes and you may not communicate anything about the contents of the drawers to the prisoners.

The riddle is thus: what is the optimal strategy for the prisoners and how many the prisoners may be saved from execution using this strategy? The solution is after the break.

Continue reading A different prisoner’s dilemma

Recent Interesting Science Articles (August 2013)

August has been a terrible month for my stock portfolio but a fantastic one with regards to science news. Plenty of reading up ahead:

  • The coolest (or is it hottest?) bit of science news is, of course, Elon Musk’s hyperloop proposal. This is sort of like an inter-city elevated train that runs inside a tube. Air is pumped out of the tube until is almost a vacuum. The idea is that the lack of air resistance enables the capsules to travel at a top speed of some 1,200 kmph while reducing the weight of the infrastructure required. The whole thing is driven by linear electric motors and costs is supposed to be kept low by building the track above existing highways. While this is a very science-fiction scenario, my take on this is skeptical. Cost is everything and despite the suggested optimizations, the proposed looks implausibly low to me as many others have already pointed out.
  • In the US, recent political trends point towards white people being critical of affirmative action being black people remain strongly supportive. But are the whites truly in favour of meritocracy? This article points out a survey that does indeed find that white Californian adults are in favour of university admissions policies that prioritize standardized test scores and high school academic achievements. But when white people are reminded that Asian Americans are disproportionately represented inside the University of California system, because Asian Americans tend to do much better at standardized academic tests than other groups, they tend turned around and favoured a reduced role for test scores instead. This suggests that white people are in favour of policies that they perceive will help their own group rather than the principle of meritocracy.
  • As we all know, evolution is a continuous process that is ongoing, for us as well as all other species that we share this planet with. This article suggests that as humans have prospered and changed the environment, the animals inhabiting that environment have evolved in response. Based on the observation of increasing skull sizes, various mammals species including mice, shrews, bats etc. seem to be evolving larger brains to successfully navigate this changed environment. In particular, the brains of small mammals in cities or suburbs seem larger those of the same species in rural areas.
  • This next article requires some knowledge of black hole physics. For years now, it was thought that if someone were to fall into a black hole, he would be crushed by the awesome gravitational forces involved. This is now changing. Not that the person would die of course, just that he would first be killed by the so-called “firewall” of energy at the edge of the black hole. This is a relatively new concept stemming from the understanding that having information flow out of a black hole would be incompatible with the Einsteinian idea that the event horizon is smooth. Instead it is now thought that a discontinuity in the vacuum, manifesting as a wall of energetic particles, exists just inside the boundary of the black hole exist. Please read the full article for how this idea came about and the implications it has for physics.
  • As someone who has some red-green colour blindness, this piece of news has personal significance. It’s essentially a review of a pair of expensive glasses that addresses red-green colour blindness. The lenses apparently contain filters that increase sensitivity to specific colour temperatures and there are specific pairs for specific types of colour blindness. Given the caveat that it works only with very bright light (so it mostly won’t work indoors and won’t work with computer monitors), the writer of the review pretty unequivocally states that it makes a tremendous and immediately noticeable difference to how he perceives the world. He was able to discern colours he had never noticed before and the colours of the world became richer and more saturated.
  • Finally, here’s an article about research into how using Facebook and other social networks tend to make its users unhappy. After having controlled for the observation that those who tend to use Facebook are just unhappy in the first place, it finds that people demonstrably becomes unhappier after each Facebook session after an extended period of tracking. The reasoning is that such use arouses envy. Since those who post to Facebook tend to do so about the best things in their lives, whether they are their best photos, best moments or best lines, such moments are exaggerated and not at all representative of such people’s everyday life, forming an unhealthy point of comparison between different lifestyles.

Algorithms: Design and Analysis, Part 1

So the latest Coursera course I’ve been taking and have now completed is Stanford University’s version of its Algorithms class by professor Tim Roughgarden. Apparently enough people asked about the differences between this course and the one from Princeton University that the question has now been addressed in the FAQ sections of both courses: the Princeton class focuses on implementation and practical programming with students in their first and second university years as the target audience. This Stanford version focuses on the analysis of algorithms and is usually taken by students in their third and fourth years.

I always knew going on that with a weak maths background, this course was going to be hard for me. It also didn’t help that I find myself much more interested in practical applications of algorithms than in studying and devising proofs of an algorithm’s correctness or performance claims. As such I didn’t put as much effort into this as I did for the Princeton course. Some observations:

  • The programming homework emphasizes actual implementations, with some tweaks, of the algorithms being studied. It’s a language agnostic course so you’re free to use whatever language you want. The professor just explains everything in pseudo-code while the homework provides a data set and you just have to type in the answers you get from processing the data for evaluation. I cheated. Having already covered much of the same ground in the Princeton course, which does provide complete implementations of all algorithms in Java, I simply used the Princeton library, tweaking code only as necessary to get the correct answers. For the most part, this made the programming homework very easy for me.
  • Th quiz portion was correspondingly harder for me. I found that I usually had a decent intuitive grasp of what to shoot for, but in many cases, lacked the discipline to thoroughly walk through the entire reasoning process. Especially since many of proofs requires a solid understanding of probability. The same goes for the final examination, which honestly speaking is slightly easier than the regular quizzes, but is timed and allows only a single attempt. We were allowed two attempts for each of the quizzes. I scored 32 out of 40 in the final exam. I did not attempt any of the optional and ungraded challenge questions.
  • I have to say that Professor Roughgarden’s lecture videos and study materials are very basic compared to the Princeton course. It’s just a talking head and slides, with no animations whatsoever. One major problem I had is that the audio frequently is not in sync with what is happening on the screen. For example, he would talk about something but when he refers to a mathematical formula or a snippet of code, it doesn’t show up on the whiteboard until a few seconds later, which certainly does not help comprehension. Also while he offers typed versions of the slides he uses as a download, he prefers the handwritten versions in the lecture videos, and his handwriting isn’t always perfectly legible.
  • I did appreciate the thoroughness with which he covers the proofs and I have to admit that I’m able to follow them much more closely than I did when professor Sedgewick went through them (which he didn’t always do in the Princeton courses. I still can’t say that it’s my favorite subject though.
  • As previously mentioned, the course does cover much of the same material so there was little here that I haven’t already seen before. I did appreciate learning about bloom filters. And after being frankly puzzled by the Kosaraju-Sharir algorithm in the Princeton course (easy to code, hard to intuitively understand) I really liked how professor Roughgarden’s explanations and the associated programming homework, allowed me to finally grasp it in my head.
  • I don’t believe that I ever saw any Community TAs in this course. By and large, as far as I could tell, the students did an excellent job of posing and answering questions themselves, with professor Roughgarden himself intervening only on rare occasions mostly related to technical problems with the course itself. I do find that the student body for this course appears more mature than that of the Princeton course, with few people posting frivolous questions or appearing to be completely clueless.
  • Certainly as the professor admits, the assessment system is currently the weakest part of the course. The course really needs a way to have students submit proofs and get them evaluated (which would have utterly defeated me I believe) but there’s simply no feasible way to scale anything like this. Also, the programming homework requires just an answer (usually just a string of numbers) without any of the performance and memory usage evaluations used in the Princeton course.

No surprise when I say that I prefer the Princeton course by a very large margin. But this is still an excellent, high quality course especially for those more interested in the analysis side of things. Incidentally I’ve been invited once again to serve as Community TA for the next offering of the Princeton course which should begin at the end of August.

 

Recent Interesting Science Articles (July 2013)

Quite a few of these articles this month, so here goes:

  •  Cloning animals is nothing new these days but, still, there is something symbolic about cloning one from a single drop of blood, as this article from BBC News covers.
  • We all know that bats can navigate using ultrasound, but could prey make use of this fact as a defensive measure? This article from Popsci covers how hawk moths, found in the tropics, are able to respond to the ultrasounds emitted by bats hunting them by responding with ultrasound clicks of their own. It may be useful to just startle the moths or it could be part of an active jamming system to hamper the effectiveness of the bats’ echolocation abilities.
  • Nearly a quarter of a century after its inception, a study into the crack babies phenomenon of the 1980s has finally ended. The term refers to an epidemic of babies born to mothers who were addicted to cocaine and were exposed to the substance while in-utero. Apocryphal stories at the time talked of babies with shrunken heads, poor muscle tone and troubling behavioural symptoms. Twenty-three years later this study found that there was no difference in IQ between such babies and those of a control group with no prenatal cocaine exposure. They did find that both groups had IQ scores that were markedly lower than the national average and attribute it to the effects of poverty.
  • This next article is somewhat like the one above: conclusions that are “obvious” are not always correct. Pop quiz: which areas are safer to live in, for developed countries at least: urban areas or rural areas? As it turns out, this article from CNN explains how contrary to intuition, the risk of injury or death from violent crime and accidents are more than 20% higher in the countryside in the United States than in urban areas. We don’t know exactly why yet, but there are some educated guesses. In the US, rates of firearm ownership are higher in rural areas for example, and these residents tend to drive longer distances over more dangerous roads. Plus it’s easier and faster to get to a hospital in a city than in the countryside.
  • Finally here’s a just for fun article about a couple of physics points from the Pacific Rim film in a Scientific American blog: specifically how much force is there in a rocket-assisted punch delivered by a giant robotic fist and can a giant monster with a huge wingspan fly its way into space? It’s just the first in a series of two such articles with maybe more to come so be sure to look out for more too.

Embassytown

200px-Mieville_Embassytown_2011_UK

I’ve heard of China Miéville from his work in other genres but this is, I believe, his first science-fiction novel, and the first book of his that I’ve read. Embassytown received some very impressive reviews so I delved into it with plenty of anticipation and not a little bit of trepidation. As it turned out, the novel is less ambitious and more traditional than I expected so I needn’t have worried myself.

As its title implies, the book is about a city called Embassytown, the sole human settlement on the alien planet of Arieka. In the universe created by Miéville, humans traverse interstellar space by going through something called the immer, while hinting that some aliens, called exots, have wholly incompatible forms of FTL technology. It so happens that Arieka is located just about at the edge of explored space in the immer. This means that while Embassytown is formally a colony of a human empire known as Bremen, it is de facto semi-autonomous due to its remote location so visits from spacecraft crossing interstellar distances, called voidcraft, are both rare and the occasion for grand celebration.

Continue reading Embassytown

The unexamined life is a life not worth living