3Where Terrorists Are Hiding Bombs
As the last, oh, thousand years of war have taught us, winning isn't necessarily as straightforward as just blowing up all the other guy's tanks and airports. In battlegrounds such as Afghanistan, the military has to deal with guerrilla tactics, where the enemy hides within the general population. It's hard to fight somebody if you don't know who or where they are until something explodes. So how do you predict the behavior of people so desperate and/or crazy that they're willing to strap bombs to themselves if necessary? Once again, with the power of math.
The University of Maryland developed software called SCARE (Spatio-Cultural Abductive Reasoning Engine) that has been consistently able to predict the locations of guerrillas' bomb caches to within half a mile, which isn't perfect, but is infinitely better than the previous "throwing darts at a map" strategy the military was forced to employ.
"It's in one of these buildings. Go."
Because just as we saw with pop music and criminal gang scenes, hiding inside what seems like random chaos are some pretty reliable parameters. For one thing, terrorists are not likely to attack too close to home (they don't want to give away the location of their base of operations, and they're not stupid), but due to physical constraints, they're also not likely to roam too far. So right off the bat, mathematicians have a pretty good idea of what area they need to look at. But that's just the start.
The software also takes into account cultural factors such as territorial allegiances. Since different guerrilla groups may not exactly be BFFs outside of their mutual hatred of a foreign military, the areas where insurgents are active tend to be closely linked by culture or religion with the area where they keep their bomb factories.
"It's not much, but if it accidentally explodes ... who gives a crap, right?"
But this kind of thing isn't just useful for tracking down terrorists. The list of potential applications includes tracking down operating bases for criminals on home soil, as well as medical diagnosis, monitoring political organizations, and presumably tracking down the home of that dog that keeps shitting in your yard.
2How New Drugs Will Affect People
One reason why it takes so long for new drugs to hit the market is that you have to test the shit out of them -- for instance, Viagra was created in 1989, but didn't come to market until almost a decade later. You just can't predict how a drug is going to react with the insanely complex pile of chemical reactions that make up the human body; you can only test it and hope for the best. What cures cancer in mice may give a human accidental superpowers. You just don't know until you try.
But imagine if we could take everything your body does, from regulating blood to fighting infections and more, and turn all of those into math equations. Theoretically, we could run human trials on new drugs without actually having to dig up volunteers willing find out firsthand if the new flu vaccine causes them to start pissing peanut butter.
"Damn you, Flintstones vitamins."
Mathematician David Eddy has laboriously gathered all the human medical data that he could and created a theoretical math person that he calls Archimedes. Of course, he needed to see whether it would react like a real human being, so he recreated a real trial for a diabetes drug. When the results came out, it was revealed that Eddy's program, a soulless black box full of hidden equations, was able to predict two of the principal findings from the study exactly. Which means that he was literally able to reduce an expensive and risky seven-year trial to numbers and bypass the whole "feed people drugs and see what happens to them" process.
And in case you're protesting that people can't be replaced by math so easily, Eddy didn't forget the human element. His Archimedes program took into account factors such as age and blood pressure, as well as fickle human flaws, such as the tendency for people to forget their pills once in a while.
Of course, this doesn't mean that we'll be replacing human trials with computers anytime soon, because putting our health in the hands of HAL 9000 is still kind of frightening, and the equation doesn't work well enough to completely replace tests on real people. But, as it stands, Eddy's work is at the very least a supplementary tool for researchers to cut down the cost and risk factor of expensive human trials. And if the occasional person still develops Hulk powers from a routine trial, then more power to them.
"BONER FEEL AMAZING -- MUST SMASH!"