This one is really lovely. Samuel Arbesman, a computational sociologist at Harvard, has created the “Milky Way Transit Authority” — a London-tube-like map of our galaxy.

As he describes it:

Our galaxy is unimaginably vast, and we really have no idea what is out there. We are discovering new planets in other star systems all the time, learning new facts about the galactic core, and even learning about whole new portions of the galaxy. This map is an attempt to approach our galaxy with a bit more familiarity than usual and get people thinking about long-term possibilities in outer space. Hopefully it can provide as a useful shorthand for our place in the Milky Way, the ‘important’ sights, and make inconceivable distances a bit less daunting. And while convenient interstellar travel is nothing more than a murky dream, and might always be that way, there is power in creating tools for beginning to wrap our minds around the interconnections of our galactic neighborhood.

I have attempted to actually make this map as accurate as possible, where each line corresponds to an arm of our galaxy, and the stations are actual places in their proper locations. However, I am not an astronomer or astrophysicist, so there are certainly inaccuracies, gaps, and room for improvement. If you have a suggestion, comment, or criticism, please to do not hesitate to contact me by emailing arbesman at gmail dot com.

(Thanks to Shareable.net for this one!)

Back in the late 90s, many newspapers reported this apocryphal exchange between Microsoft CEO Bill Gates and General Motors:

At a recent COMDEX, Bill Gates reportedly compared the computer industry with the auto industry and stated: “If GM had kept up the technology like the computer industry has, we would all be driving twenty-five dollar cars that got 1,000 miles per gallon.”

Recently General Motors addressed this comment by releasing the statement: “Yes, but would you want your car to crash twice a day?”

I was abruptly put in mind of this old joke when reading the latest news about Toyota’s crash-prone cars, because some of the fatal problems now appear to be based not in the mechanics of the cars, but in their software. The New York Times today reported on the story of 77-year-old Guadalupe Alberto, who died when her 2005 Camry accelerated out of control and crashed into a tree; “the crash is now being looked at as a possible example of problems with the electronic system that controls the throttle and engine speed in Toyotas.”

The point is, cars have gradually employed more and more software as control systems, to the point where, as the Times notes …

The electronic systems in modern cars and trucks — under new scrutiny as regulators continue to raise concerns about Toyota vehicles — are packed with up to 100 million lines of computer code, more than in some jet fighters.

“It would be easy to say the modern car is a computer on wheels, but it’s more like 30 or more computers on wheels,” said Bruce Emaus, the chairman of SAE International’s embedded software standards committee.

Maybe we shouldn’t be surprised if Toyota winds up wrestling with bug-caused crashes. Once software grows really huge, its creators are often unable to vouchsafe that it’s bug-free — that it’ll work as intended in all situations. Automakers have a vested and capitalistic interest in making sure their cars don’t crash, so I’m sure they’re pretty careful. But it’s practically a law of nature that when code gets huge, bugs multiply; the software becomes such a sprawling ecosystem that no single person can ever visualize how it works and what might go wrong. Worse, it’s even harder to guarantee a system’s beahvior when it’s in the hands of millions of users, all behaving in idiosyncratic ways. They are the infinite monkeys bashing at the keyboard, and if there’s a bug in there somewhere, they’ll uncover it — and, if they do so while travelling 50 miles an hour, possibly kill themselves.

The problems of automobile software remind me of the problems I saw two years ago while writing about voting-machine software for the New York Times Magazine. As with cars, you’ve got software that is performing mission-critical work — executing democracy! — and it’s in the hands of millions of users doing all sorts of weird, unanticipated stuff (like double- or triple-touching touchscreens that are only designed for single-touch). Let’s leave aside the heated question of whether a manufacturer or hacker could throw an election by tampering with the software. The point is, even without recourse to that sort of skulduggery, what I found is that the machines so frequently crash, bug out, or just do head-scratchingly weird stuff that it’s no wonder so many people refuse to trust them.

So what’s the solution? Well, in the world of election software, many have suggested open-sourcing the code. If thousands of programmers worldwide could scrutinize voting-machine software, they’d find more bugs than the small number of programmers currently working in trade secret. And theoretically this could improve public confidence.

Would the same process improve automobile code? Should the software in our cars be open-sourced?

» MORE...

You may have heard about You Are Not A Gadget: A Manifesto, the new book by Jaron Lanier, the inventor of virtual reality. It’s been getting quite a lot of praise; in the New York Times, Michiko Kakutani called it “lucid, powerful and persuasive”, and the New Yorker suggested that “his argument will make intuitive sense.”

I was recently assigned to review the book by Bookforum, and that review is now on the newsstands. It’s an omnibus analysis of three books, loosely bungee-corded together by their common concern for how the Internet is changing the way we think, write and communicate. I had been looking forward to Lanier’s book in part because I really enjoyed his provocative 2006 essay “Digital Maoism”, and more importantly because I’ve interviewed him once or twice and thought he was an incredibly thoughtful guy. (What’s more, he’s a musician, and I often find I like the way musicians think.)

Unfortunately, You Are Not A Gadget is pretty dreadful. It has flashes of absolute genius, but is ruined by a flaw common to most woe-is-the-digital-age books: They begin by claiming the Internet is filled chiefly with either completely idiotic culture or nasty, meanspirited, anonymous commentary — but they offer almost no evidence to document this. So you spend the entire book listening to Lanier explain why the Internet is such a dreadful place, without ever being convinced that it actually is. (I am currently sketching out an upcoming blog entry where I rant about this literary trend at numbing length.)

In contrast, the other two books I reviewed were written by academics who wondered about the impact of the Internet on society, but who actually decided to collect some data on it; not coincidentally, they came up with conclusions considerably less apocalyptic, and considerably more nuanced, than Lanier.

Research! It’s not just for breakfast any more.

Anyway, the review is online here for free at the Bookforum site, and a copy is archived below!

Floating Signifiers
The Internet hasn’t killed the English language — yet
by Clive Thompson

In the late 1870s, the advent of the telephone created a curious social question: What was the proper way to greet someone at the beginning of a call?

The first telephones were always “on” and connected pairwise to each other, so you didn’t need to dial a number to attract the attention of the person on the other end; you just picked up the handset and shouted something into it. But what?

Alexander Graham Bell argued that “Ahoy!” was best, since it had traditionally been used for hailing ships. But Thomas Edison, who was creating a competing telephone system for Western Union, proposed a different greeting: “Hello!,” a variation on “Halloo!,” a holler historically used to summon hounds during a hunt. As we know, Edison — aided by the hefty marketing budget of Western Union — won that battle, and hello became the routine way to begin a phone conversation.

Yet here’s the thing: For decades, hello was enormously controversial. That’s because prephone guardians of correct usage regarded it (and halloo) as vulgar. These late-nineteenth-century Emily Posts urged people not to use the word, and the dispute carried on until the 1940s. By the ’60s and ’70s, though, hello was fully domesticated, and people moved on to even more scandalously casual phrasings like hi and hey. Today, hello can actually sound slightly formal.

» MORE...

Life isn’t easy for the “scaly-foot gastropod”. This humble snail lives in hydrothermal vent fields two miles deep in the Indian ocean, and is surrounded by vicious predators. For example, there’s the “cone snail”, which stabs at its victims with a harpoon-style tooth as a precursor to injecting them with paralyzing venom. Then there’s the Brachyuran crab, which has been known to squeeze its prey for three days in an attempt to kill it. Yowsa.

Ah, but the scaly-foot gastropod has its own tricks. To fight back, it long ago evolved a particularly cool defense structure: It takes the grains of iron sulfide floating in the water around it and incorporates it into the outer layer of its shell. It it thus an “iron-plated snail”.

Oh yes way. Scientists discovered Crysomallon squamiferum in 1999, but they didn’t know a whole lot about the properties of its shell until this month, when a team led by MIT scientists decided to study it carefully. The team did a pile of spectroscopic and microscopic measurements of the shell, poked at it with a nanoindentor, and built a computer model of its properties to simulate how well it would hold up under various predator attacks.

The upshot, as they write in their paper (PDF here), is that the shell is “unlike any other known natural or synthetic engineered armor.” Part of its ability to resist damage seems to be the way the shell deforms when it’s struck: It produces cracks that dissipate the force of the blow, and nanoparticles that injure whatever is attacking it:

Within the indent region, consolidation of the granular structure is observed within and around the indent. Localized microfractures exhibit tortuous, branched, and noncontinuous pathways, as well as jagged crack fronts resulting from separation of granules, all of which are beneficial for energy dissipation and preventing catastrophic brittle fracture. Such microfracture modes may serve as a sacrificial mechanism. Upon indentation, inelastic deformation will be localized in the softer organic material between the granule interfaces, which allows for intergranular displacement and friction while simultaneously being compressed down into the softer ML. Shear of iron sulfide nanoparticles against the indenter surface is expected, in particular since penetrating attacks take place off-angle rather than directly on top of the shell apex, and can be facilitated by intergranular displacements during yielding of the OL. This provides a potential grinding abrasion and wear mechanism to deform and blunt the indenter (since biological penetrating threats are in reality deformable as well) that will continue throughout the entire indentation process.

Beyond awesome. This is Darwinian evolution mixed with, like, Burning Man.

Being scientists of biomimicry, the authors surmise that if it were possible to reverse-engineer the entire shell — it’s not just the outer iron layer that’s cool; there are also two inner layers with gooey nougat that are equally important in defending the snail — they could produce superstrong materials for military defense and “load-bearing”.

Fair enough. But personally I’m satisfied just to have more pure science that proves, yet again, the inexhaustible Weirdness Of The Briny Deep.

Iron snails, people! Iron snails.

(Thanks to the Eco Tone blog for alerting me to this one!)

Back in 1997, chess grandmaster Garry Kasparov played against the IBM supercomputer Deep Blue, and lost. At the time it was widely regarded as a huge victory for artificial intelligence. But as Kasparov points out — in a fantastic new essay about computer chess in the New York Review of Books — experts had long predicted that a computer would eventually beat a human at chess.

That’s because chess software doesn’t need to analyze the game the way a human does. It just needs to do a “brute force” attack: It calculates all the possible games several moves out, finds the one that’s most advantageous to itself, and makes that play. Human grandmasters don’t work that way. They do not necessarily “see” the game several moves out. Indeed, they can’t — as Kasparov points out, chess is so complex that “a player looking eight moves ahead [faces] as many possible games as there are stars in the galaxy.”

Instead, people who are truly great at chess use the peculiarly human qualities of understanding, insight and intuition: They study oodles of games, encode that knowledge deeply in their brains, and practice incessantly. “As for how many moves ahead a grandmaster sees,” as Kasparov concludes, the real answer is: “Just one, the best one.” We like to think that artificial intelligence is replicating human smarts, but in reality it does something quite different. One doesn’t learn much about human intelligence by examining the way computers play chess, just as one doesn’t learn much about computer intelligence by examining the way humans play chess. They are fundamentally dissimilar processes.

But this gave Kasparov a fascinating idea. What if, instead of playing against one another, a computer and a human played together — as part of a team? If humans and computers think in very different ways, perhaps they’d be complementary. So in 1998 Kasparov put together an event called “Advanced Chess”, where humans played against one another, but each was allowed to also use a PC with the best available chess software. The chess players would enter positions into the machine, see what the computer thought was a good move, and use this to inform their own human analysis of the board. Cyborg chess!

The results? As Kasparov writes:

Lured by the substantial prize money, several groups of strong grandmasters working with several computers at the same time entered the competition. At first, the results seemed predictable. The teams of human plus machine dominated even the strongest computers. The chess machine Hydra, which is a chess-specific supercomputer like Deep Blue, was no match for a strong human player using a relatively weak laptop. Human strategic guidance combined with the tactical acuity of a computer was overwhelming.

The surprise came at the conclusion of the event. The winner was revealed to be not a grandmaster with a state-of-the-art PC but a pair of amateur American chess players using three computers at the same time. Their skill at manipulating and “coaching” their computers to look very deeply into positions effectively counteracted the superior chess understanding of their grandmaster opponents and the greater computational power of other participants. Weak human + machine + better process was superior to a strong computer alone and, more remarkably, superior to a strong human + machine + inferior process.

This stuff really fascinates me, because so much of our everyday lives now transpire in precisely this fashion: We work as cyborgs, using machine intelligence to augment our human smarts. Google amplifies our ability to find information, or even to remember it (I often use it to resolve “tip of the tongue” moments — i.e. to locate the name of a person or concept I know but can’t quite put my finger on). Social-networking software gives us an ESP-level awareness of what’s going on in the lives of people we care about. Tools like Mint help us spot invisible patterns in how we’re spending, or blowing, our hard-earned cash. None of these tools replace human intelligence, or even work the way that human intelligence works. Indeed, they’re often cognitively quite alien processes — which is precisely why they can be so unsettling to some people, and why we’re still sort of figuring out how, and when, to use them. The arguments that currently rage about the social impact of Facebook and Google are, in a sense, arguments about what sort of cyborgs we want — or don’t want — to be.

What I love about Kasparov’s algorithm — “Weak human + machine + better process was superior to a strong computer alone and … superior to a strong human + machine + inferior process” — is that it suggests serious rewards accrue to those who figure out the best way to use thought-enhancing software. (Or rather, those who figure out a way that’s best for them; people always use tools in slightly different, idiosyncratic ways.) The process matters as much as the software itself. How often do you check it? When do you trust the help it’s offering, and when do you ignore it?

(That photo above is by Elke Wetzig, and released for use under the GNU Free Documentation License!)

My latest column for Wired magazine is now online, and it’s a fun topic: I analyze the downside of becoming Twitter famous. You can read the full text below — or for free at Wired’s site, or in print if you race out to a newsstand this very instant and pick up a copy! — but the gist of the argument is simple: If you have too many followers, the conversational and observational qualities that originally make Twitter fun start to break down … and you’re left with old-fashioned (and often quite dull) broadcasting.

If you like the column, go check out Anil Dash’s excellent blog post “Life on the List”. He describes what it’s like being put on Twitter’s “Suggested User” list — i.e. the list that Twitter publishes that recommends interesting people to follow. Those who are put on the list quickly begin amassing thousands of followers a day, which is precisely what happened to Anil, who now has over 327,000 people following him. And Anil encountered precisely the phenomenon I described: He didn’t like it, and thinks it partly ruined his experience of Twitter.

Anyway, read on! (That picture above, by the way, is the lovely illustration by Helen Yentus and Jason Booher that appears in the print copy of Wired.)

In Praise of Obscurity
by Clive Thompson

When it comes to your social network, bigger is better — or so we’re told. The more followers and friends you have, the more awesome and important you are. That’s why you see so much oohing and aahing over people with a million Twitter followers.

But lately I’ve been thinking about the downside of having a huge online audience. When you go from having a few hundred Twitter followers to ten thousand, something unexpected happens: Social networking starts to break down.

» MORE...

How much TV do you watch? What’s the highest level of educational level you’ve attained?

According to data gathered by the web site hunch, these two aspects of your life are “almost perfectly inversely correlated”: The more advanced your degree, the less time you spend staring at the tube.

Here’s the background: Hunch is web site that gives you customized recommendations based on you answering questions about what you do and don’t like. (After parsing my replies to several questions, it advised me against buying an Apple tablet, for example.) In theory, the more people use Hunch, the more Hunch knows about our preferences and the smarter its recommendations get. So to gather even more information about people’s preferences more quickly yet, the site has a section called “Tell Hunch About You”, where you can answer oodles of survey-like questions about your demographics, your likes, dislikes, habits, patterns of consumption, beliefs, etc.

Over 66,000 people have answered questions about both their educational level and the amount of TV they watch. When the Hunch folks assembled the numbers, here’s what it looked like, according to their blog:

It turns out that increasing educational level is almost perfectly inversely correlated with daily TV consumption. Of the 22% of Hunchers who completed no more than a high school education, only about 12% of them watch no TV but a full 25% watch 4 hours or more each day. On the other end of the spectrum, of the 26% of Hunchers who have completed at least a PhD, about 17% of them watch no TV and only about 16% watch 4 hours or more each day.

Here’s another way to look at the data. For 3 the groups of Hunchers who completed no more than 2 years of college, about half of each group watches 2 hours or more of TV each day. But that’s true of just 44% of those with a 4 year degree, 37% of those with a masters, and 35% of those with a PhD or higher.

If you follow the links they have some nice graphs to illustrate the numbers. Of course, assuming these data hold up (anyone know how well it compares to other studies of the same thing?), it’s still an open question as to which is the chicken and which the egg. Does having higher education make TV somehow less attractive as an activity? Or are people who already don’t like TV more likely to pursue higher education?

Me, I stopped watching TV in college, largely because when I moved out of home and into a rooming house in dowtown Toronto, I didn’t own a TV — and there was no common room in my building, i.e. no dorm-like place where everyone would hang out and watch. Since you often solidify your leisure-time habits in your teens and early twenties, I never resumed watching TV regularly even after I left college. (The sheer metric tonnage of stuff I’ve missed is kind of amazing: I’ve only seen about three episodes of The Simpsons, for example.) It’s not that I don’t like TV; I actually love it when I do see it, in part because the quality of TV has become so spectacularly high in the last decade. These days, I usually follow one show at a time — Mad Men right now — so I watch an hour a week. (However, in a somewhat cosmic irony, I actually married a TV critic, so I also catch lots of snippets of different shows while she’s reviewing them.)

The point is, while I can correlate going-to-college with stopping-watching-TV, the relationship seems in my case to be entirely circumstantial. I don’t think there’s anything in my educational makeup per se that would turn me away from TV — at least not the sort high-quality dramatic or comedic stuff.

What about you guys?

(By the way, Hunch has so much data on its users that it has found all manner of other fascinating correlations, which they’re reporting on the Hunch blog. Check out their profile of “birthers”, what your video game system says about you, and what Hunch users think about global warming.)

(Thanks to Caterina Fake’s tweet for alerting me to this one, and to dailyinvention’s Creative-Commons-licensed photostream for the picture above!)

Last week, Sean Patrick Fannon got an interesting idea on how to raise money for Haitian relief. Fannon works for RPGNow, a web site that allows tabletop RPG creators to upload their games as print-on-paper PDFs, set a price, and sell them via download. Fannon emailed all the game creators who sell through his site and pitched them this concept: If they’d donate a game, Fannon would bundle them into a single $20 downloadable purchase, sell as many as he could, and donate the proceeds to Haitian relief. Pretty soon dozens of game designers were uploading their gaming items — ranging from little monographs on Second World War munitions to entire 300-page book-length game manuals that would cost $45 if you bought them in printed format. So many designers offered their wares that the bundle now contains $1,481.31 worth of product … which you can get for only $20.

A pretty awesome value, eh? Indeed, it’s excellent enough that RPGNow has raised, as of today, a remarkable $132,325.00 for Haiti. (That total includes direct $5-to-$10 donations made through the site, too.)

It’s a cool enough story on its own. But there’s also some interesting economic behavior here, too, on the part of the game designers. On the one hand, they’re giving copies of their stuff for charity — i.e. forgoing possible profits. Or are they? Since there’s no additional costs in making more PDFs, the question of foregone profits hinges entirely on whether the creators think the folks buying the $20 pack might otherwise be prospective customers. As Greg Stolze — a game designer who donated some of his own work to the project, and who alerted me to this sale — pointed out in an email, the project …

… highlights the plasticity of an idea’s value in an internet market, that’s for damn sure. I stuck my book eCollapse in the bundle: It hadn’t been much of a mover, so I don’t think I’ve lost even hundreds of dollars of sales by throwing it in, and it’s probably the same with almost everyone else. No one’s really taking a serious hit because we don’t have to risk sunk material costs, just abstract potential profits.

It puts me in mind of Chris Anderson’s argument in Free, which is that when the cost of something goes to zero, it evokes new economic phenomena: Consumers become more experimental, and creators can focus on the free (or near-free) mass distribution of their works, while making money off other stuff — like add-on services and goods, customization, or the like. In this case, the game designers can feel proud that they’re helping raise a lot of money for a good cause while also possibly expanding the universe of people who know their work, and might be likely to pay for new works in the future.

Stolze, I should point out, pioneered one of my favorite new economic models that leverages digital-age behavior: The “ransom” model of publishing. Back in 2005, he announced on his blog a concept for a new game he was designing, and told his audience that he was accepting donations for it. If he reached $600 in pledges, he’d design the game and release it as a free PDF for anyone to download. In other words, if enough of his hard-core fans decided they were willing to pay for the game, anyone could get it. It thought it was a brilliant concept, and as it turns out …

… it’s so brilliant that it’s roughly the model behind Kickstarter, a web site that launched last year offering precisely the same service: Anyone can announce a project, set a financial goal, and see if enough people are willing to support it! Fittingly, Stolze is using Kickstarter to pursue the ransom model again — this time to publish short fiction. Check his page out!

And if you’re into RPGs, seriously, go buy that $20 pack for Haiti. It’s a crazy value for the money.

I’ve got nothing to say about this, really; it’s just awesome.

(Thanks to TYWKIWDBI for this one!)

Wolfram Alpha is a super cool question-answering system. Ask it about something factual, and it’ll offer up whatever specific info it has — such as the dimensions of a #10 screw or a definition of “20/50 vision” (including an eye chart fuzzed out at the right line!) Wolfram Alpha can also answer queries that require it to collect together, parse and compare bits of data, such as finding the “10 nearest stars” or comparing the populations of Corsica, Sardinia and Sicily.

But what happens when you ask it a metaphysical question? I tried the query above — “Does God exist?” — and cracked up at the answer:

I’m sorry, but a poor computational knowledge engine, no matter how powerful, is not capable of providing a simple answer to that question.

Good to know Skynet’s on board with the non-overlapping magisteria, eh? Interestingly, this is not a stock answer that the engine kicks out whenever it cannot parse a question. (Actually, the stock answer seems to be “Wolfram|Alpha isn’t sure how to compute an answer from your input” — which is what you get when you ask, for example, “Why is Nickelback so awful?”) No, the God question was clearly anticipated by the Wolfram people, who inserted this nice little easter egg. (No pun intended. No epistemological allegory intended?)

It also turns out Wolfram Alpha has a number of other great easter eggs. For example, if you ask it “How many roads must a man walk down before you can call him a man?” the result is pretty awesome. (I got this latter example from Mashable, which compiled a list of the best 10 they could find.)

I also enjoyed typing in my name — Clive Thompson — at which point Wolfram Alpha assumed I was asking about the town of Clive in Iowa and the town of Thompson in New York; it described their relative populations, plotted them on a map, and calculated that it would take 1 hour and 50 minutes to fly from “Clive” to “Thompson.”

This humidifier is awesome for two reasons.

The first is that it uses no electricity: Instead, water is drawn up through the wood and evaporated into the air. According to the designers — the Japanese firm Okada — this process evaporates water six times faster than if you left in in a glass. At that pace, it’d be quite effective.

The second reason, of course, is how unbelievably gorgeous it is. Okada calls the device the “Mast”; it’s supposed to look like a sailboat. Google’s translation of the Japanese site selling the Mast is pretty mutilated, but it’s poetically mutilated, so I’ll use it:

Like a yacht sail (mast) the wind, the natural moisture to dry air liberality, also allows them to subtle and refreshing fragrance of cypress. Its appearance, we thought a yacht floating in the cool water and soft drinks will also give a visual sense.

How to use: Please put the water on the boat portion of the mast.

Absorbs water and sails made from scrap parts of the Seeds of cypress, cypress and give a faint fragrance of Hodoyoi natural moisture.

I want one! They’re only $70 US, though I bet the cost of shipping would be brutal.

(Thanks to the Design Less Better blog for finding this one! They’re the ones who called it the “zen humidifier”.)

While doing some egosurfing today, I hit upon this interview I did last spring with WNYC’s On The Media. It was a segment on “the rise of indie video-game designers”, so most of the segment is various game folks talking about how the online world — and the downloadable markets on the consoles — are bringing a new spirit of innovation to gaming.

In the final minute of the broadcast, though, we speculated on an interesting question: What’ll it be like 30 years from now when US has a president who grew up playing video games?

BROOKE GLADSTONE: Well, let me ask you the big cultural question here. If, in fact, video games may become the largest form of home entertainment in America within the next year, in terms of dollars spent, how will that change the culture?

CLIVE THOMPSON: I certainly would hope that this breeds — particularly if we get more inventive forms of games — a type of a cultural mindset that’s more interested in the complexity of the world and complex systems, because the one thing that you do when you play a video game is you sit down in a state of total ignorance. You don’t know what you’re supposed to do. You don’t know how this game works.

And the process of figuring out what to do is really what is fun about the game. In fact, people often stop playing the game once they’ve figured it out. And that’s a great inquisitive mindset that video games quite uniquely tap into.

So, you know, 30 years from now, imagine a president who did this, you know, as their primary cultural activity for 25 years.

I don’t know if this is actually true. I like to think that game-playing — of all sorts, ranging from Uncharted to backgammon — leaves you with some useful mental habits, but I know far less about what’s genuinely useful in a presidential mindset. But either way, we’re eventually going to find out: A 20-year-old today who runs for president in 2040 will have spent his or her entire life playing games, I’ll bet.

Here’s a study with an interesting finding: If you want to get better results on Google, try using a shorter query.

I found this while doing research for a story about automated “question answering” systems. I was reading through the work of James Allan, a computer scientist at the University of Massachusetts, and read his paper “A Case for Shorter Queries, and Helping Users Create Them” (PDF here). In it, he and his coauthor Giridhar Kumaran conducted an experiment: They took the query Define Argentine and British international relations and ran it through a search engine. (They don’t specify which one they used.) Then they ran various similar queries that used fewer words — “sub queries” — such as define britain international argentina or define britain relate argentina. Each time, he graded the relevance the search engine’s results, expressed as their “average precision” on a scale of zero to 1.0.

So which sub-query produced the best results? The shortest one. It was only two words long — britain argentina — but it scored 0.626, quite a lot better than the original, full-sentence query, which scored only 0.424.

Why would short queries work better than longer ones? Possibly because they contain fewer “noise terms” — common words like define or and — which might muddy the search results. Human language is filled with ambiguity; one of the big challenges for a machine is taking a human question and figuring out what, semantically, it’s actually asking. In that sense, using fewer words would reduce the number of potential ways the machine can misunderstand you.

Except the truly strange thing in that example above is the question was asking about British and Argentinian international relations — yet the best results came from removing the words “international” and “relations”. I’d have expected those to be important words, no? But that’s precisely the point Allan is getting at here:

Sub-queries a human would consider as an incomplete expression of information need sometimes performed better than the original query.

This suggests, of course, that the best way to get results on a search engine is to radically strip your query down even further than you think is useful. Or maybe start with a regular query, and if you don’t like the results, try making it shorter and shorter.

Then again, it’s hard to know if this would really work. I’m not privy to what’s going on behind the hood of most search engines today. Allan’s paper discusses several ways for question-answering systems to have the computer automatically shorten a query before feeding it into the knowledge database; but his paper is a few years old, so maybe these techniques are already common amongst search engines — maybe they already reformat our queries into semantically shorter formats.

What do you guys think? Anecdotally, have you found that super-short queries work better than longer, sentence-like ones?

Why do people eat unhealthy, high-calorie fast food? Is it because they don’t realize how bad it is for them — or do they realize it and just don’t care?

Two years ago, New York City bet it was the former. The city government passed a law requiring all fast-food chains to display the calorie count beside their food listings, right next to the prices. The theory was that if you forced people to see how many calories they were consuming, they’d make healthier picks. But did it work?

Apparently so. A group of Stanford University professors got Starbucks — one of the chains that began listing calorie counts — to give them the records for 100 million transactions at stores in New York, Boston and Philadelphia, over a 14 month period. Boston and Philadelphia do not have a similar law in place, so they were the control group. Sure enough, when the scientists compared consumption patterns at Starbucks in the three cities, New York’s law appeared to have had a small but significant impact: People made lower-calorie orders. As their paper — “Calorie Posting in Chain Restaurants” (PDF here) — concludes:

We find that mandatory calorie posting does influence consumer behavior at Starbucks, causing average calories per transaction to decrease by 6% (from 247 to 232 calories per transaction). The effects are long lasting: the calorie reduction in NYC persists for the entire period of our data, which extends 10 months after the calorie posting commenced.

Here’s what really intrigued me, though: At Starbucks, people behaved differently with food than with beverages.

When they calorie postings went up, customers quickly switched to lower-calorie Starbucks foods. But they didn’t switch to lower-calorie beverages. The scientists found calorie intake from food dropped by 14%, while the drop in calories for beverages was “negligible.” Indeed, the majority of the overall calorie reduction — that 6% drop — consisted of customers simply deciding not to order any food while at Starbucks. But when it came to drinks? They barely changed at all.

This is rather surprising, because some of those Starbucks drinks are insanely high-calorie. Say you order a Venti “Caramel Brulee Creme” with nonfat milk? That’s 480 calories, 70 of which are fat. Or how about a Venti “Double Chocolaty Chip Frappucino Blended Creme” with whipped creme? Friend, you just inhaled a whopping 670 calories, 200 of which were pure fat.

Given how fattening these drinks are — and, frankly, how nauseating they are; I have no idea how anyone can gag back these syrupy cocktails — why wouldn’t customers, when presented with the calorie count, pick something healthier? Some of these things are like pouring rendered tallow directly down your throat. (I’m actually not kidding: If you ate a half stick of butter you still wouldn’t come close to the calorie count of Venti “Pumpkin Spice Frappucino Blended Creme.”) Who drinks these revolting concoctions? I mean, “Mint Chocolaty Chip Frappuccino blended creme with Whipped Cream”? Even the whipped cream has to be chocolate-flavored? These customers do realize that there’s probably a corner store only a block away selling actual, regular, normal coffee for like 50 cents, right?

Okay, I’ll calm down now. The question is, why did the calorie information change the way Starbucks customers ate — but not stuff they drank? The researchers don’t speculate. But I think it’s because of the psychology of Starbucks itself. People who go to Starbucks are motivated primarily by the desire for a drink, not a piece of food. So even if they’re forced to confront the calorie count on their favorite, hideously repellent bucket of candied snot (sorry, I can’t stop myself here), they’re probably not going to change their mind. They’ll just decide to forgo the chocolate graham crackers.

Indeed, if you want further proof of this, consider another fascinating bit of data the scientists discovered: Starbucks customers actually overestimate the amount of calories in their favorite drinks. Before the enforced nutritional-posting law came into effect, the scientists polled Starbucks customers and asked them to guess how many calories were in their beverages. Amazingly, fully three-quarters of customers overestimated the amount. They thought the drinks contained 90 calories more than in reality. So that helps further explain why the labeling didn’t shift the customers’ drinking habits: They already knew the drinks were bad for them, and they didn’t care. Heck, some of them were probably relieved to discover they were drinking a mere 680 calories, instead of 770. Woo!

I should point out that the New York Times ran a story today that suggests the calorie-labeling law may prove to be a dud in the long run. Apparently the trend line for Starbucks customers went in a U-shape: After the law came into effect, their calorie intake dipped significantly, but then began creeping up again, and holiday binge-eating brought it entirely back up in line with Philadephia and Boston. Still, I have to say I’m impressed that New York managed to get a 6% reduction at all.

Maybe they should just hire me to stand outside various Starbucks locations and hector people. Yeah, that’d work.

(The photo above of that totally gross Peppermint Mocha Twist drink comes via stephenccwu’s Creative-Commons-Licensed Flickr photostream!)

James Pennebaker is a psychologist who specializes in analyzing how your use of seemingly innocuous words — like “I”, “we”, “he” or “she” — reflects your emotional state. He’s used his technique to study everything from the relationships of top Al-Qaeda members to the complexity of presidential candidates’ speech patterns. I first heard about his stuff because I tried out AnalyzeWords, a web app Pennebaker and his colleagues created that studies your recent tweets and deduces your “emotional style,” “social style” and “thinking style”. (Apparently, according to my recent tweets I’m “depressed”, “personable”, and “in-the-moment.”)

Anyway, I started reading through Pennebaker’s academic papers and hit upon this really fascinating one: The effect of testosterone on word usage.

Pennebaker knew that many studies have linked increased testosterone levels to aggression, negative moods, increased sex drive, and even things like improved
spatial skills and impaired verbal ability. So that got him wondering: Does higher testosterone affect how we use language?

To test this, he found two subjects who were undergoing testosterone-injection therapy: A 60-year-old man who was getting the injections to restore his upper-body strength, and a 28-year-old biological woman who three years into a transgender program to become a man. Pennebaker got writing samples — like email and journal entries — from the subjects before and after the injections.

The results? As Pennebaker wrote in a summary of the study:

Overall, testosterone had the effect of suppressing the participants’ use of non-I pronouns. That is, as testosterone levels dropped in the weeks after the hormone injections, the participants began making more references to other humans … One function of testosterone, then, may be to steer people’s interests away from other people as social beings.

On the one hand, that matches up with traditional claims about increased testosterone — that it makes you into a kind of gently sociopathic, type-A jerk. But the really interesting thing is that Pennebaker did not detect any other big changes in the subjects’ emotional states. “Contrary to commonly held beliefs,” he writes, “changes in testosterone levels were unrelated to linguistic markers of mood state, aggression, sexuality, achievement, and references to perceptual or cognitive processes.” So maybe some of the stereotypes about testosterone need to be revised.

Of course, it remains to be seen if Pennebaker’s findings hold up; with only two subjects, the results are very provisional. If you want to read his paper — “Testosterone as a Social Inhibitor: Two Case Studies of the Effect of
Testosterone Treatment on Language” — it’s online free here.