Sunday, February 12, 2006

Franchise-based Online Mapping

There are few experiences in life more stressful than navigating an unfamiliar city. Services such as Google Maps have made this somewhat easier by combining beautiful user interfaces with simple directions. All these services are based on data from a few mapping companies whose employees drive every road in America each year, making note of lane mergers, stop signs, and dead-end streets, all correlated with differential GPS accurate to inches. The services that use this data give directions that are resultantly almost military in precision: "Drive 1.1 miles, turn right on Martin Luther King Expressway, go .9 miles" etc. This would be a fine state of affairs if we conducted all our unfamiliar driving in the dead of night, with 20/15 vision and a zen-like focus on the green and white signs that validate the directions Google et al. see fit to provide. Unfortunately that is almost never the case, and human beings recognize as much.

Human-provided directions tend toward including landmarks, and are more likely to include visual descriptions (especially, research has shown, directions given by females). But while a resident of your new city might know that there's a giant red warehouse just before the road you need, Google isn't likely to. If only there were a way to proxy for recognizable landscape features... as a matter of fact there is. Lamentable or not, cities of any size are now packed full of nationally popular chain stores and restaurants. Happily, the most stressful places to drive -- high traffic thoroughfares -- are also the most likely to attract franchises with consistent and recognizable signage. Instead of counting distances along a highway to the tenth of a mile (which is unlikely to be useful to anyone this side of Von Neumann), Google could tell you to take a left two blocks past the third McDonald's, then turn right after the Holiday Inn. The businesses themselves are architecturally distinct, and their signs are elevated dozens of feet above the roadway, blasted with stadium-grade electric lighting, and made orders of magnitude larger than the street signs harried travelers would otherwise have to search for.

These national chains are also likely to have consistent nomenclature, which makes mapping, say, all the McDonald'ses, Wendy'ses, and Burger Kings in the nation a trivial affair for a company like Google, which already has copious amounts of directory information. In fact there is already a Google Maps Mashup that helps you locate the nearest fast-food restaurant. Factored into plain-language directions, the chain-store uiformity that we decry for monotonizing the American landscape could make the American landscape much easier to navigate. Sure this isn't going to help you pick one street in a residential neighborhood (unless your neighborhood is lucky enough to feature a McDonald's!). But using street names to navigate is much easier when you're driving 20 miles per hour through the garden district than when you're driving at 40 miles per hour down an eight-lane boulevard. Just a thought - listening, Google?

Saturday, August 06, 2005

Ha ha! Go ahead and steal my idea, GE! I've got your number.

Version: PGP 8.1 - not licensed for commercial use:


Sunday, July 31, 2005

On Haunted Houses

I have an older SUV that lacks much of the padding and upholstry buyers expect nowadays. My radio is also kaput, and driving home one day this spurred an interesting idea: everybody's familiar with that "hairs stand up on the back of your neck" feeling you get from time to time. It is strongly correlated with dark, creepy places and a history of horror movie fandom. But what makes a place creepy? Bloody handprints certainly help, but I'm talking about the subconscious lizard-brain feelings of unease we get walking around in abandoned houses and mental institutions - and I think I have at least a partial answer. It seems so obvious in retrospect that I assume I am not the first person to make this suggestion, though Googling yields nothing.

Feelings of creepiness are an evolutionary response to the auditory sensation of movement we can't see a cause for. Consider your average Sci-Fi channel Real Ghost Hunters special. Teams of "researchers" with thermal imaging equipment and audio gear load up and camp out in what is usually a dilapidated residence. Rarely do they visit, say, 80s-era corporate office buildings (ostensibly because fewer evil deeds of a magnitude to create hauntings take place there). Anyway, the thermal imaging equipment is brought along because feelings of a ghostly presence often coincide with unexplainable cold spots that can be electronically verified - giving some validation to people's very subjective feelings that they aren't alone. My thesis is that it is the cold spots that cause these feelings of unease, even if people can't feel the change in temperature.

In well-constructed buildings that are relatively small, and in the great out of doors, the air temperature tends to be homogeneous. Ambient noises in such surroundings - the sound of the wind in the trees or of one's own breathing - travel in predictable billiards-table paths and reflect at billiards-table angles. Not all environments are so homogeneous, however. Submariners and deep-sea fishermen are aware of the thermocline, a sharp boundary in sea temperature that sonically divides the sea less than, say, 150 feet deep, from that 175 feet and below. This temperature boundary layer is nearly opaque to acoustic signals. The temperature differences in "haunted" houses aren't nearly so sharp - but the density difference in the air is enough to lense sound away from its obvious source even though we can't see that anything has changed. Draftiness is a stereotype in stories about haunted buildings, and I think it is very likely a piece of unexamined folk wisdom.

So let's set this up, then. Let's say that in the mid-afternoon I assemble with my friends to tough out a night in a putatively haunted house. A few of us have analog watches, but all of us are breathing, so we have a foundation of small ambient noises that none of us consciously notices. As the sun sets and the air outside the house cools, air that was still in an uniform-temperature environment is now subject to mixing. Cooler air filters through cracks in the ceiling, just after night falls, and a column of cool air develops in the center of the room. The warm air/cool air boundary reflects some noises back to us much sooner (and therefore louder) than if they had travelled across the room unimpeded. Some sounds we are used to hearing, on a subconscious level, disappear as they are refracted or blocked by the air lens. In a homogeneous environment (such as we are evolved to) such a disturbance could have been created by a sneaking lion, say, or an approaching horde of attacking pygmies. If we turned a torch on either of those two, we could verify the causes of our unease. In this drafty house, however, the feeling is exacerbated by the fact that there is no visible cause for the change in the atmosphere. The evolutionary distress signals we are accustomed to follow have marooned us here, sensing movement (because the air is constantly shifting, and the apparent noise source shifts accordingly) where there is none. Freaky stuff.

The funny thing is that if these "haunted" houses were occupied, background noise from televisions, conversations, and the washing machine would probably be sufficient to drown out the lens effect. It is when one is relatively quiet (but not absolutely quiet, in which case there would be no noise to refract) that the lens effect becomes apparent.

There are two experiments that could verify this: send two teams of volunteers into a building reputed to be haunted, and record their pulse and breathing rates as they spend the night. One team gets noise-cancelling headphones, the other doesn't. I'd bet a considerable amount that those with the headphones would have a cakewalk (deaf volunteers would work even better, potentially).

The other is one you can do yourself. The next time you have some spare time, and a willing volunteer, blindfold yourself and have someone quietly light a candle at some distance behind you. Ask the person to snuff out the candle at his convenience but without telling you when. You'll be amazed to see how you can "hear" a burning candle in a quiet room, and tell when it's extinguished.

Which brings me to the question: do we become less afraid of ghosts and monsters as we grow older because we learn better, or because we hear worse?

Sunday, July 24, 2005

How to Start a Revolution with SMS

Ukraine's Orange Revolution was an excellent example of how a large and sufficiently dedicated group of people can, without the use of violence, force political change. The Orange Revolution, like most revolutions associated with mass protests, had an identifiable cause: popular unrest at the notion that the incumbent party had contrived to steal the presidential election.

Being a revolutionary in a repressive state is tough work: leaders with the charisma, motivation, and following to become opponents of the ruling party inevitably attract the attention of the state as soon as, if not before, they become popular with the population at large. It is tough to maintain cohesion within an opposition movement, or to trust collaborators to follow through when the time for action (even peaceful action) nears. For these reasons, oppressive governments have persisted even as their ability to censor information effectively has vanished with the advent of the internet and anonymizing technologies like Tor (which I am using to post this entry!)

Another technology poses a much greater threat to tyrannical regimes, however. Cell phone use has become ubiquitous in newly developed countries. In China it has leapfrogged landline telephony as a means to connect the population without requiring the extensive investment associated with conventional telephone networks. Some economists have argued that the spread of cell phone technology will be the single greatest factor in increasing the standard of living in developing countries.

Consider the following scenario, however:

A committed group of tech-savvy Zimbabwean expatriates decides it can no longer watch as President Mugabe's regime bulldozes slums. The bulldozing program constitutes an atrocity, certainly - but there is no single act that can get the population out of their houses and onto the streets in the way the stolen election motivated the Orange Revolution Ukrainians. So these tech-savvy expatriates elect to invent one. Large crowds, they know, have a psychology all their own - ask anyone who has ever been involved in a soccer riot. Even without a specific casus belli, they know that if they can get enough people out on the streets, the chanting and jeering will start and the closing bell will have tolled for the Mugabe regime.

As Paris Hilton discovered, cell phone networks are far from impregnable to electronic attack. The Expats hack into the phone records of Zimbabwe's cell phone companies, and download all the phone records from Harare, the capital. This is where they get clever. They crunch the data to establish social networks from the call logs. If I call 555-5555 three times a day, an observer can assume that the number belongs to a friend or family member of mine. I probably have him in my address book. We'll call him John. When the analysis is complete, the Expats write a script to send one of a couple of dozen possible messages to each of the people I call frequently, all of them ostensibly sent from my number. The text differs slightly (and would be contrived so that no single person got the same message from two different people), but the essence is "come quick - there's a rally to oppose mugabe" or simply "you'll never believe this - come to XXX square, quick." With the messages staggered somewhat so that the rate seems natural, the following occurs:

Seeing a plausible and interesting text message coming from my number, John decides to call me and find out what's going on. Because everyone in Harare has just had this same idea, the cell phone network is busy. Perhaps our Expats have had an additional hand in this. Aware that heavy call volume can cause cell networks to go down, John decides to see what's going on, and starts to leave his house - indeed, he sees people marching toward the square even now.

Inevitably, some people will get a message from someone who is 1. physically present next to them or 2. someone from whom such a suggestion could not plausibly have come (say, a cautious mother). Even in these cases, the people who are taken in, and who begin streaming toward the centers of power, will constitute a pretty strong indication that something is Really Going On.

The trick to all this is that, unlike a traditional flashpoint, this one will come without warning. By the time the ruling party realizes it has an angry mob on its hands, the mob will number in the hundreds of thousands, and mobilizing troops or police to stop them will be impossible. Unlike an organic revolution, this revolution will involve no inside conspiracy to be infiltrated by spies or to be nipped in the bud at the first sign of unrest. It will happen so quickly that government-controlled media will be too late to broadcast warnings against it, and it will all be over by the time anyone - even the protestors themselves - understands what's happening. And I rather think it will happen bloodlessly.

What's more, it all rests on a technology no modern government can deny its citizens. Even Syria is seeing skyrocketing cell phone use among its population; governments that don't allow their citizens to go wireless risk lagging even further behind economically (which is unacceptable, because governments must consider outside threats as much as inside ones.)

Of course, this scenario is not in the least limited to Zimbabwe - any country with an unpopular and undemocratic government, and a wired (well, wireless) and youthful population will be equally susceptible. Nor do the hackers involved have to be expats. Hacktivism has a short but storied history, and while socially motivated hackers have never pulled off a revolution, they have accomplished feats that require no more technical skill than this would involve. The revolutionary heroes of the next wave of democratization may remain forever anonymous, having planned, organized, and executed a revolution in Harare or Tehran, from the comfort of their homes in Moscow, Madrid or Skokie -- or from their offices in the Pentagon. After all -- the idea of starting a revolution by SMS isn't exactly a secret.


I am not sure that this blog will ever reach a wide readership. I don't really seek one; really it's an outlet for me to get off my chest those ideas I have that are potentially important enough to be burdensome, but not potentially profitable enough (in a financial sense) to expend much effort on. Some of them are of a nature that demands anonymity on my part, so this format is very helpful.

On a more personal note: I am not really sure yet that I'm not a crank. A lot of my motivation for starting this blog is that I hope that people wiser and more experienced than I can tell me whether I'm full of shit. I'm ok with being full of shit; I'd just like to know one way or the other, so I can stop bothering my friends and family members with impractical or impossible ideas. Thus: fire away, please. You're doing me a favor.