The Wayback Machine - https://web.archive.org/web/20110510101552/http://arvindn.livejournal.com/
Arvind Narayanan's journal [entries|archive|friends|userinfo]

Pinboard.in as a Lightweight Database [Apr. 25th, 2011|11:42 pm]
[Tags|, , ]

Summary.

  • A tagged/annotated list of web links with some rudimentary programming is surprisingly useful for many simple content curation and presentation tasks.
  • Pinboard.in is particularly well-suited as a store for such links due to its reliability and nice API.

I noticed some of my academic peers collecting lists of press mentions of their work, and I wanted to make such a list for myself. I realized that the only way I'd maintain such a page, and not eventually abandon it, was if it was dead simple to add or delete a link. The ideal UI would be to tag an article via a bookmarklet and have it automatically show up on the list.

So I quickly whipped up a script to do that; you can see the result here.

Not so delicious. The first version of the script used delicious as the backend, which gave me headaches right away. One of the links wasn't showing up for some mysterious reason, and after banging my head for a while, I realized it was delicious's API that was buggy. Worse, transiently buggy.

That's when I decided it was finally time to quit delicious (for everything, not just this project) and jump ship to the site that all the cool kids were talking about—pinboard.in. Best ten bucks I ever spent. The API is compatible with delicious, so I didn't even need to change my code.

Now that I had the script in place, I kept finding all sorts of new uses for it. I'd been meaning to collect my writing and software into lists for years; now that I had the right tool, it just happened. I also convinced my collaborator on DoNotTrack.Us to use my script for his bibliography.

Why does this work? The difference in usability in each case was so dramatic that I figured something was going on here that was worth thinking about. This is my best attempt to explain the combination of factors that make this approach so appealing:

  1. Maintaining a list of links is the kind of thing that happens via numerous tiny efforts spread out over time. If I had to log in to edit a document each time I wanted to add a paper to a bibliography, or change a minor detail like published to unpublished, the administrative overhead would overwhelm the actual work. Use of a bookmarklet here is a form of in-place editing.
  2. The bookmarklet automates a lot, including the title and the link itself; the backend script takes care of automatic chronological sorting. The latter, after all, is the trivial feature that led to the explosion of blogging.

Finally, here's a totally different application for which I found pinboard perfectly suited as a "database": making collage posters, which is a recent hobby. Here is my latest one: 

 
Click to embiggen (note: ginormous size)

I tag the images on pinboard, and my collage script uses the tags to guide the layout. The alternative would have been way more cumbersome. 

Parting thoughts. First, I'm curious to hear about any other potential use cases. One application I'm already considering is spaced repetition learning—bookmark an article and it will be injected into your RSS reader (or other reading list) at spaced intervals.

Second, I would urge developers of websites and CMS software to consider incorporating this feature. Not necessarily using pinboard, but the broader concept of creating/editing content via a bookmarklet. The primary challenge would probably be in communicating the workflow to users who are not familiar with it.

Link2 comments | Leave a comment

Unbearable Incongruity [Mar. 29th, 2011|07:50 pm]
[Tags|, ]

My walks frequently take me past Facebook's offices—they're just a few blocks from where I live. College terrace is a thoroughly residential neighborhood that I can only describe as a rustic paradise. Standing in the midst of these cute-as-a-button homes and lush vegetation, where time seems to have come to a stop and the only sound is that of crickets chirping, I stare transfixed at the innocuous building that doesn't bear a corporate logo or any signage at all for that matter, except for the street number "1601." My brain refuses to grasp the fact that lurking inside is a fiercely competitive corporate titan dead set on a path to world domination.
LinkLeave a comment

Observations from Chennai [Mar. 9th, 2011|03:43 am]
[Tags|, , , , ]

I see snapshots of my hometown, Chennai, once every 2-3 years. The changes are always stunning. 

Economy 

A sustained period of 8-10% inflation-adjusted GDP growth has extraordinary effects on a country. For one, rapid economic growth has a habit of bulldozing ‘culture’ and ‘tradition’ like a mosquito. If you know me you know I consider that to be largely a good thing. More to the point, affluence seems to be eroding what I consider the bad aspects of tradition in Chennai, such as superstition, but not the good, such as Carnatic music. 

Living through rapid societal change is weird, even if it is for the better. The generation gap is ridiculous. And it’s a confusing time for everyone. The country is playing catch-up with decades, even centuries of progress in the developed world all at once. 

I think several factors have contributed to the pace of Westernization. It seems to me—forgive me for being an armchair economist for a second—that high GDP growth coupled with high inflation makes foreign goods cheaper to buy compared to local ones, and that when people import material products, they will inevitably import a bit of the culture that produced them. Then there is of course the information revolution which is making the world smaller. 

In a way, India got the IT revolution before the industrial revolution. Everyone has a mobile phone, but mechanization is still minimal to nonexistent in many areas of life, such as food—whether agriculture or cooking in the kitchen. Many daily activities still involve the use of pre-electric era technology.

On that note, signs of leapfrogging are everywhere. People have laptops but they never had desktops. Mobile phones without land phones. And now I’m hearing of a crazy plan to put high-efficiency solar units with advanced battery storage in villages, eliminating the need for the electric grid altogether. That’s a 100-year leap. 

By the way, the 8-10% growth figure I mentioned is only an average; for city dwellers who’ve worked hard, incomes have increased even more than that, perhaps averaging 15%. The improvement in the economic status of my social group compared to two decades ago is just astounding.

Government and the masses 

The wealth gap has probably increased—I haven’t checked the stats—but at least in the cities, the trickle-down is in full swing. Cars are replacing motor-bikes and motor-bikes are replacing bicycles. Wages for minimally-skilled and unskilled labor, such as driving and housework, have gone up a lot, although there is still a long, long way to go.

But the progress is encouraging. People can live with dignity, even in the slums. No one is starving. Caste matters less each year, and the crime rate, already low, continues downward. I saw a much lower level of general unrest. I didn’t think this day would come so soon, but I saw a happy people. 

Government has had a big role to play. The water shortages and rolling power blackouts are gone. Panhandling is way down; apparently there’s been a rehabilitation drive, and it’s clearly working. There are many areas that need to improve a lot, primarily traffic and pollution/cleanliness, but there are encouraging signs. 

The current approach to road traffic congestion seems to be to build flyovers (overpasses) everywhere. This is clearly not sustainable, but the city metro rail (MRTS) is being slowly rolled out, and it remains to be seen if it will make a serious difference.

As for pollution—for those who’ve never been, the city, like most Indian cities, is basically overflowing with garbage, and the air is noxious from vehicle exhaust as well—the near-term outlook is not very rosy. One interesting Government initiative in this area is a big push toward public art. It is easy to pooh-pooh this, but I am wondering if it will make a (small) difference by mending the broken window, thus reducing litter. 

One area where Government effort is simply inadequate is the expansion of the city. Cities are the future of the country—there is nothing to friggin’ do in the villages. Agriculture needs to be mechanized and the entire population needs to pour into megacities as quickly as possible.

While Chennai is expanding “rapidly” by first-world standards, it’s not nearly fast enough. There is a huge demand for labor in the cities, and rural people are dying—sometimes literally—to get there, but the bottleneck is infrastructure development and (de-)regulation. 

Unfortunately I think this is less a matter of Government inefficiency, although that plays a big role, as Government policy. Our honchos are still set in their socialist-era ways of thinking. I find that depressing. How anyone with any brains can fail to realize that rapid urbanization is vital is beyond me. 

Food

The restaurant scene has improved dramatically. A newly affluent middle-class is demanding an end to the idly monopoly, and the city finally, finally has an adequate number of restaurants serving other cuisines. I would have been miserable if I hadn’t been able to get Chinese food regularly.

Foreign cuisines are heavily Indianized, of course, but so what? Most restaurants in America are Americanized as well. I should note though that the majority of (say) Chinese restaurants in America are run by Chinese immigrants, but that’s not the case here. For this reason, we have far more ‘multi-cuisine’ restaurants than those serving a particular foreign cuisine.

Food remains cheap—an entree at a nice sit-down restaurant is $1-$3—but prices are going up quickly. Road-side food remains dirt cheap, but then you’re likely to get actual dirt with it.

There are a few successful chains, but US names have had trouble making an impact. Amusingly, ghetto US chains like KFC set themselves up as exotic sort-of-upscale dine-ins over here. 

In one of the more obvious and predictable signs of affluence, coffee shop culture has arrived with a bang. Again the successful chains are local, Cafe Coffee Day being by far the most prominent. No surprise—it is nearly impossible for a chain to enter India without tweaking the model heavily. 

One coffee shop I went to had a designated make-out area. I mean, it’s not marked as such, but everyone understands that that’s what it’s for. It’s brilliant if you consider the prevalent cultural factors, and has become my favorite example of tailoring businesses to local conditions.

Fashion

I will be blunt—it is hard to overstate how badly people dress in Chennai.

The older generations are the worst. For this group there is little distinction between formal and casual wear. For men, it consists of half-sleeve “dress” shirts, slacks, and sandals. I will leave you to picture that abomination for a minute :-)

For women, traditional attire begins and ends with sarees. Ugh, sarees. They are so ridiculously uncomfortable that I can’t help thinking they must have been invented by men to keep women in the kitchen. At any rate, there is no reason to keep wearing them now that textile technology has advanced beyond the ability to make rectangular pieces of cloth.

Things are somewhat better among the youth. Men’s wear is gradually reaching parity with the West (with some oddities: polo shirts are called “t-shirts”; actual t-shirts aren’t worn much). Women’s wear shows no such inclination. Thankfully, though, sarees have been replaced by salwar kameez. Perhaps in one more generation, social norms will have relaxed enough to allow women to be proud of their bodies rather than having to hide them with baggy attire.

My complaint is more than that. People seem generally oblivious to what they’re wearing and whether it looks good. When I was growing up I never thought about it, because I hadn’t been anywhere else much and because fashion doesn’t matter when everyone is poor. But you’d think affluence would change things. It hasn’t.

Physical culture is also essentially nonexistent. A few young men seem to be taking up weight training, but by and large physical activity is simply not seen as part of everyday life. The vast majority of women, in particular, get no exercise whatsoever (other than walking and housework, if those even count).

Assorted differences from America

The big things are of course all different, but it’s relatively easy to get used to them. It’s the little things that make you pay attention to the differences between Indian and American life and go, “huh.” Here’s a random selection of examples.

Intellectual property laws haven’t caught up, and are poorly enforced. You can find stores carrying internationally recognized brands such as PUMAA and “Converge All Star”. It's not as extreme as China, though. For the record, I'm opposed to most forms of intellectual property in a normative sense; trademark is one that I'm marginally in favor of.

The huge differences in relative costs can result in some pretty bizarre situations. Buying a microwave oven or a chapati maker (tortilla press) can be a major event. Get this—there’s a salesperson who visits you at home to give a demonstration. It only adds a little bit to the cost of the machine, and it seems to be necessary because people are still so gingery about mechanizing their kitchen. I bet millions of housewives are silently terrified about becoming unwanted if cooking gets easier.

My haircuts cost Rs. 70, which is 5x what they did a decade ago, but still only about 6% of what I pay in the US. Insane.

Surprisingly, there are still hardly any foreigners around. On the plus side there are now a large number of immigrants from other states, and they seem to have a monopoly of new businesses in certain sectors, such as restaurants.

There still isn't, and it is doubtful if there will ever be, a notion of privacy or personal boundaries similar to the West. An example: I was shopping for shirts and asked a salesperson for help locating my size. I said I was a 42, to which he responded, “Are you sure? I think you’re a 44.” The guy next to him—remember, stores are teeming with salespeople because labor is absurdly cheap—chimes in: “no, I think he’s right, he’s a 42. Look, he doesn’t have a big belly or anything.” I want to emphasize that this is a perfectly normal conversation to have in India.

OK, I’m done. Is there anything in all these changes that I consider to be a negative? As someone who considers modernity good per se, I have to say no. My only quibble is that some things aren’t changing fast enough. For example, there are still manual rickshaws, although they are very rare. I’m looking forward to seeing what the next decade will bring.

Link16 comments | Leave a comment

Nexus S: First Impressions [Mar. 2nd, 2011|10:06 pm]
[Tags|, , , ]

I've had my new Nexus S for a week. I prefer not to use Apple products for a variety of reasons, and I'm a fan of Android, so it was the obvious choice. It's a big leap from my old phone, a G1.
 
To set the context for this review, I'm not using the phone for development and have no plans to hack it or jailbreak it. I don't know what CPU it uses or how much memory it has, and don't care. I'm just a regular of my phone, although I may use it for more tasks than most. While my opinions about technology and business ethics influence my choice of product, they have no further bearing on this review.

Weight. The nicest thing about the phone is its weight—it's almost as light as a Nokia dumbphone, and way lighter than the iPhone 4. I don't think weight matters any more from a utilitarian perspective (as long as it weighs less than a brick), but then phones are partly fashion accessories these days.

Notification light. There isn't one. The iPhone doesn't have one either. I just don't get it! Could someone explain it to me what people do without a notification light, because the mystery of how iPhone owners cope is driving me crazy. Are y'all totally dependent on vibration and ringtones? Or do you just take a long time to get back to people?

Fortunately for the Nexus S (unlike the iPhone) the screen is SuperAMOLED, which apparently means that black pixels use no power! What I'm getting at, if you haven't already figured it out, is that the notification light can be emulated on the screen itself in software, and the app noled does just that. 

Based on the fact that it took people a while to figure out the noled strategy, I'm pretty sure Samsung wasn't really thinking ahead when they decided to remove the notification light and were merely aping the iPhone. Ugh.

Keyboard. Speaking of aping the iPhone, the Nexus S has no physical keyboard either. But unlike the notification light I've decided to be happy with it, because it seems like on-screen keyboards are improving rapidly, and while still not nearly as good as a physical keyboard, they will eventually get there and even surpass them. In fact it seems clear that physical keyboards will soon be extinct, so the sooner I get used to on-screen typing the better.

Polish. Android has come a long way in terms of polish (it always had the lead in terms of features). There are far fewer annoyances with Gingerbread, the latest version. I'm talking about little things like automatically detecting when I remove the phone from my ear (based on the accelerometer) and turning the screen back on. There's still a long way to go, but since OS updates for the Nexus S are not at the mercy of my carrier and come directly from Google (right?) I'm looking forward to regular improvements.

Battery life is still a huge problem. Spoiled non-Android users have no idea how bad it is. I'm thrilled if it lasts a whole day and I don't mind plugging it in every night. But even that is touch-and-go. If I listen to music it won't last the whole day, otherwise it usually does. (For comparison, the G1 would almost never last a whole day even with fairly light activity, and if I tried to listen to music it would last less than an hour.)
 
This is all assuming some background process doesn't go berserk, which happens often. I've been using the Advanced Task Killer app to kill everything when it looks like the phone is running hot or battery is running out too quickly, which seems to improve things. (Depressingly, we still don't seem to be at the point where the benefits of multitasking outweigh the problems, although I think we'll get there soon.)
 
Apps and security. A few things like weather and alarm that should be considered essential phone features don't work very well and have made me look for alternatives on the Market. In fact what I need isn't an 'alarm' app; it's a 'sleep' app that should allow me to silence the phone, play brown noise, and set an alarm all from one screen. This kind of user-centric design is unfortunately not Android's forté.
 
The selection of apps on the Market is quite adequate. I don't recall ever thinking, "I wish that were available for Android." The big problem with apps is security & privacy, which I worry about more than most people. The granular permissions notification is supposed to be a great thing in theory but doesn't work well in practice. Many apps ask for all permissions ("just in case"), and even when they don't, it's unclear to the user what the app actually needs and whether the request is legitimate. Perhaps the biggest drawback is that it's a take-it-or-leave-it deal for the user.

Worse, there's now (fairly serious) malware on the Market. These got taken down, but there will be more. This isn't something that the permissions dialog even theoretically helps with. What I've been doing is largely sticking to apps that I've heard about and have some reason to trust.
 
Concluding thoughts. Convergence marches on. I often don't know if I'm talking to someone via text or IM, and don't care. Facebook and Android can now sync contacts. I find myself using Yelp on the phone even when I'm at my computer because the phone already knows my location and because I can click on phone numbers. My phone has also become my book-reading device (I previously used the iPod touch due to battery issues on the G1), and I'm glad the number of things I carry with me at all times has dropped from 4 to 3 (phone, keys, wallet).
 
Speaking of phone-keys-wallet, it is sad but entirely predictable that phones haven't yet taken over payment and authentication. Maybe NFC will make a difference. Fingers crossed.
 
In summary, my phone finally does "basic" smartphone functions like email, books and yes, phone calls well enough that I mostly just forget about it. The most exciting apps are those that interface with the real world in some interesting way—Shazam, Goggles, Wakemate and Car Locator to name a few. The best part is that this space is still in its infancy. Fun times.
Link2 comments | Leave a comment

Balwadis in Chennai: a First Impression [Feb. 21st, 2011|03:01 am]
[Tags|, , , ]

On a recent trip to Chennai, my aunt who volunteers for ASHA took me along on a trip to 4-5 Balwadis (government-run pre-schools, apparently supplemented by private contributions). Here's a writeup of some notes I took. This was originally meant as feedback for ASHA but I decided to post it here as well.


What made the most striking impression upon me was that in every school I visited, the kids were curious and friendly. They were not crying or fighting, took to strangers easily, and most imporantly, seemed eager to learn and explore. I refer not only to classroom instruction—I gave a two-year old a pen and piece of paper to see if she could draw, and after scribbling for a few seconds she began to investigate how the pen worked (it was a retractable ballpoint), and quickly figured out how to operate the spring mechanism. The general impression I got was that with the right care and teaching, these children will go very far.

As for the Balwadis themselves, the conditions are very basic, to put it mildly. They are overcrowded and understaffed (if I understand correctly, there are enough 'aayas' (caretakers) but not many qualified teachers). Many of the schools did not have a working gas line for cooking mid-day meals, and some didn't even have running water. The classrooms are kept relatively clean (compared to the surrounding slums). The children sit on the floor and write on slates with chalk. The walls are filled with brightly colored instructional materials such as pictures of birds and animals with their names.

My recurring observation was that a few changes that would cost little or nothing could make a real difference. One big area is hygiene—while it is not terrible, there is a lot of room for improvement. One of the aayas was chopping vegetables on the floor when we visited; I assume they all do this. A raised countertop (which was available) would have been much more suitable. A child who repeatedly indicated that she needed to go to the bathroom was ignored until she gave up and urinated in the classroom; this was not cleaned properly.

Another area that could use improvement is study materials. Take the use of slates and chalk: I know this is how things were done traditionally, but pencils and notebooks are cheap these days. It would allow the teacher to monitor the child's progress, but more importantly, taking the notebooks home would encourage parents to keep their kids in school by showing them that the children are learning and making progress over the months and years. There was also a conspicuous absence of toys. Young children learn by playing, and not just by receiving verbal instruction. Toys would have the added benefit of encouraging children to go to school.

Needless to say, I'm unfamiliar with the workings and the constraints of the system and my suggestions must be taken with a pinch of salt. In summary, I saw promise and hope, but there need to be systemic changes in order to realize the potential, and I'm glad that ASHA is involved in funding and volunteering. I wonder if it would be possible to develop a blueprint consisting of ideas and procedures that have shown results in individual schools that can then be distributed to all the teachers.
LinkLeave a comment

American Nerd [Jan. 24th, 2011|11:46 pm]
[Tags|, , , ]

There isn’t a single native-born American among the world’s top 250 chess players. The number for women is worse: none in the top 900 or so.[1]

Chess is by no means the only pursuit that Americans largely choose to ignore. As a computer scientist working in America, I know all too well that my field of research and many others that are stereotyped as ‘nerdy’ are utterly dominated by immigrants, and that the immigrant to native-born ratio for women is even more skewed. What is different about chess is that the stark results are clearly measurable.

Many fear that this severe disdain for mathematical/technical/scientific fields means that the loss of America’s competitive edge is imminent. I find these arguments unfounded — there is no reason why the supply of immigrants will dry up in the foreseeable future.[2] The only bottleneck is the bizarre policy of tightly limiting visas and turning away highly qualified workers.

If I’m not raising the economic argument, why am I talking about this? For two reasons: the first is that I find the stigmatization of nerdy behavior only slightly less morally repugnant than racism or sexism. I consider it a significant but largely unremarked blemish in the social progress that the country has made. The second reason is that I’m curious to know why America is uniquely and extraordinarily bad in this respect. I’d welcome any thoughts on this question.

[1] I’m using these two lists. Larry Christiansen is the top native-born American currently active, and Elizabeth Vicary the top female. And to be clear, I am comparing women with women. If you compound the paucity of Americans with the paucity of women, incomprehensible numbers result: not a single native-born American female among the top 25,000 players.
[2] While I recognize that there is room for different views, the economic question isn't the point of this post and I don’t feel like arguing about it. I merely wanted to state my opinion.
Link9 comments | Leave a comment

A fateful dodgeball game [Jan. 24th, 2011|01:46 am]
[Tags|, , ]

Some time in '95-96, when I was about 14, I was in Sivakasi participating in a chess tournament. Yes, I played tournament chess back in those days, and aspired to be a 'professional' when I grew up, with little understanding of what the word meant.

Sivakasi is an utterly godforsaken town known primarily for being the locus of India's firecracker industry, and was once notorious as a hotbed of child labor. There was absolutely nothing to do there. We could get a whole plate of idlis for Rs. 7 (about fifteen cents) but even a fat kid can eat only so many idlis. Which is why a few of us young chessheads found ourselves playing dodgeball outside the tournament hall one afternoon in spite of the scorching heat.

We were mean. We only let other kids into the game if we thought they were going to be chess stars, like we clearly were. I did have my claim to fame, and had no trouble getting into the group, but I'll save that for another post. Far more interesting is the story of one of the kids we wouldn't let in, an eight-year old girl called Humpy Koneru.

She was clearly uncool, because she was coached by her dad instead of a professional coach. Even though she'd recently won the national under-8 championship, it was just a fluke. I mean, jeez, under 8. Besides, her dad had made ridiculous statements to the press, such as claiming she was a child prodigy. And finally, her name was Humpy. No, there was going to be no letting her into that dodgeball game.

Humpy Koneru said a big fuck you to all of us and went on to be the youngest woman to ever become a grandmaster (at 15), and the second highest rated female in history.

Another kid in that dodgeball game, Magesh Chandran, also went on to become a grandmaster.

As for me, I dropped out of chess the following year.

There isn't really a moral or even a point to this post. Yes, I was mean, and that was wrong, but so are most 14-year olds. I have many tales from my chess-playing days, and I simply want to record them for posterity. If nothing else, I will read these posts when I'm old and half-senile, and smile at the bittersweet memories.

Link1 comment | Leave a comment

Is Chess with Queen Odds a Provable Win? [Jan. 15th, 2011|09:57 pm]
[Tags|, , ]

What is the result of a perfect chess game?

That is perhaps the best-known example of a mathematical question that is well-defined and in principle soluble in a finite number of steps, and yet not only beyond the abilities of what our computers can currently calculate, but is likely to remain so forever, at least via any known algorithmic approach.

Where machines fall short, we still have human intuition and experience. Most chess experts believe that chess is a draw, but probably wouldn't bet their life on it. Some Masters have expressed support for the possibility that chess might be a win for White. But at the very least, almost everyone would agree that chess is not a win for Black — for that would mean that the opening position is a Zugzwang — and I for one would bet my life on it (not that I expect the question to be decided within my lifetime).

So we have a hypothesis that is a virtual certainty, namely that chess is either a draw or a win for White, but for which we have absolutely no clue as to proof techniques. Of course, there is no shortage of examples of this from the realm of the infinite, such as the twin prime conjecture, but there is something even more unsatisfying about finite problems that are undecidable in practice.

What if we make things easier for the machine? It is obvious to a rank beginner that a perfect game with a rook handicap is a win for the side with the material advantage. No, make it a queen! Surely that must be a provable win?

Not so fast. Even against a crushing asymmetry in material, it is not too hard to avoid mate for a couple of dozen moves, which means that calculating all the way to the end of the game is beyond the reach of search-based algorithms.

However, it is possible that completely different methods, perhaps more in the realm of theorem-proving than search, or a hybrid of the two, might be successful. As far as I know, there is no serious research program to investigate questions of this nature. And I'd like to spend the rest of this post talking about why that is.

In many ways, the much-publicized and celebrated victory of IBM's Deep Blue over Kasparov in 1997 was acutally a tragedy for game-playing research. The popular perception of the result was that chess had in some sense been cracked by machines; game research lost its glamor, and funding dried up. In reality, all that had happened was that chess programs had gotten really good at searching game trees and that computers had gotten faster.

Indeed, if you start from a closed position where strategy dominates and tactics are of relatively little use, top human players can still trounce computers. (Of course, the machines are programmed not to get into such positions in the first place in the course of a regular game.) The point is that chess-playing programs in no way duplicate human abilities at chess, and there is a lot of scope for further research.

Moving on, Kasparov points out the intriguing and hugely counterintuitive results of games between human–machine teams, and laments that this avenue has not been explored further. Yet another area largely ignored by game researchers is games like arimaa, similar in broad respects to chess but specifically constructed to be hard for machines. Sure enough, there is a vast difference between human and machine abilities, and in fact, humans have been getting stronger at arimaa relative to computers!

In short, it is only a minor exaggeration to say that game-playing research died just as it was about to get interesting.

I have hopes that there will be a revival one day. Perhaps as "general intelligence" inevitably becomes more sophisticated, chess programs that don't do any explicit search will come into vogue as a showcase for generic reasoning ability. What I hope I've done in this post is to expose a potentially new problem domain — proving wins in easily won positions — which, if pursued seriously, might lead to progress in underdeveloped areas of machine intelligence.
Link3 comments | Leave a comment

2010 in review [Jan. 6th, 2011|04:04 pm]
[Tags|, , ]

Best movie I watched: Inception. I've written about the dream-state being much more mysterious (and probably more powerful) than most people think, and being more similar to the wake-state than we are comfortable admitting. So in addition to being a load of fun, the movie had a deeper meaning for me.

Runner ups: The Social Network, Inglourious Basterds, Waiting for Superman. Many of my Silicon Valley acquaintances are upset about the portrayal of nerds in The Social Network, but I thought it was spot on.

Best book I read: The Checklist Manifesto. Mind-blowing and life-changing.

Runners-up: The Innovator's Dilemma, The White Tiger.

Best app: Rdio. Subscription music is absolutely the only model that works for me. After Apple bought and killed Lala in May, I didn't have anything for a few months until Rdio came along. Now I'm happy again. Let's see how long this one will last.

Runner-up: F.lux. Everyone needs to use this, especially if you have any problems at all going to bed and getting up on time. The only excuse for not using it is not knowing about it. And now you know about it.
Link12 comments | Leave a comment

An analysis of cold medications [Dec. 26th, 2010|06:08 pm]
[Tags|, , ]

I find common cold medications rather interesting. There is of course no "cure," or even a way to shorten the duration of the illness. On the other hand, there are a variety of symptom-relief strategies with varying effectiveness. This has led to a plethora of drugs on the market; this page lists "154 widely used nonprescription cold remedies" (!) along with their ingredients.

I cleaned up the spelling and formatting errors in that list and wrote a script to make a histogram showing how many different products each ingredient occurs in, color-coded by category.


Ingredient vs. number of products.  
Click image to embiggen

Why did I do it? Simply because I could. It might also be because I get colds a lot and making this chart porn gives me a feeling of being sort of in control, which is an altogether different kind of placebo.

Here are some observations:
  • There are 6 major categories: antitussives, decongestants, antihistamines, analgestics/antipyretics, expectorants and local anesthetics.
  • There are about 40 different ingredients. The chart shows the ones that occur at least 3 times, which is 27 of them.
  • While I've listed "belladonna alkaloids" under "bogus" (yeah, I'm not the biggest fan of alternative medicine), I'm sure many others on the chart are equally ineffective. I don't know which ones. The studies are typically inconclusive.
  • Many ingredients have multiple effects (for example, most of the decongestants are also stimulants). I had to make guesses at which effect is relevant to cold symptom relief, although in most cases it is obvious.
  • There are an average of 2.5 ingredients per product (this is a separate analysis, not from the chart). Most have 1-4 ingredients but a couple have as many as 6.
The raw data is here. I can imagine doing other (equally useless) things with it, such as looking at which ingredients occur together. Finally, in case it needs saying, I'm not a doctor and am in no way qualified to give medical advice; this is just a pretty chart and nothing more.
Link4 comments | Leave a comment

navigation
[ viewing | most recent entries ]
[ go | earlier ]