I find this very disappointing...
For the last two decades, human cognitive superiority had a distinctive sound: the soft click of stones placed on a wooden Go board. But once again, artificial intelligence is asserting its domination over gray matter.
In February, at the Taiwan Open â€" Go's popularity in East Asia roughly compares to America's enthusiasm for golf â€" a program called MoGo beat two professionals. At an exhibition in Chicago, the Many Faces program beat another pro. The programs still had a head start, but the trend is clear.
Arrayed by opposing players trying to capture space on its lined 19x19 grid, the black and white Go stones can end a game in 10^171 possible ways â€" about 10^81 times more configurations than there are elementary particles in the known universe.
Faced with such extraordinary complexity, our brains somehow find a path, navigating the possibilities using mechanisms only dimly understood by science. Both of the programs that have recently defeated humans used variations on mathematical techniques originally developed by Manhattan Project physicists to coax order from pure randomness.
In a more recent brain-scanning study, Japanese researchers compared professional and amateur Go players as they contemplated opening- and end-stage moves. Both displayed parietal lobe activity. During the end stages, however, professionals had extremely high activity in their precuneus and cerebellum regions, where the brain integrates a sense of space with our bodies and motions.
Put another way, professionals fuse their consciousness into the decision tree of the game.
Go players have an ability "to think creatively and prune the search tree in an aesthetic sense," said Atherton. "They have a feel for the game."
Enter the Monte Carlo method, named by its Manhattan Project pioneers for the casinos where they gambled. It consists of random simulations repeated again and again until patterns and probabilities emerge: the characteristics of an atomic bomb explosion, phase states in quantum fields, the outcome of a Go game. Programs like MoGO and Many Faces simulate random games from start to finish, over and over and over again, with no concern for figuring out which of any given move is best.
"The surprising, mysterious thing to me is that these algorithms work at all," said Hearn. "It's very puzzling."
"There's a strong tendency in humans to have a conceit about how far we've advanced," said Doshay. "But we've only really started programming computers."
Originally Posted by Coma
But this doesn't do it like a human, it does it better. I don't see why they're surprised it works - no AI is needed to play something like Go, just pure logic.
Originally Posted by TestECull
When they can concoct a robot that can drive any automobile just as a human does, I'll be impressed...and downright scared too, considering the dumbass moves I've seen real humans do. There's no telling what kind of idiotic stunts a robot might pull, seeing as robots don't have fear...
Originally Posted by TestECull
No, they don't...not on public highways. Right now, robot cars are still too stupid to not hit anything unless they implant a track in the roadway for them to follow, and then they're no more complex than a 10 dollar line tracking mouse kit for electronics 1 students in high school.
The only self driving cars that don't hit things and don't follow a track are merely 1:1 scale RC's piloted by humans standing on rooftops nearby...
Originally Posted by Arkanor
If it makes you guys feel any better, computers process data way faster than we can, and the only way this can be made to work is because computers are so good at brute forcing their way through problems.
Crunching through millions of games of Go just to finish 1 game? If you lose I guess you can just say the computer's been playing way more than you, heh.
Originally Posted by dharmaBum
This is somewhat inaccurate. Human brains (and down the food chain to at least insect nervous systems) process massive amounts of information fairly quickly. Visual perception for example; the mind can take in constant, continually varying and very complicated environmental input and make almost instant sense of it. Or, if you prefer in tech-geek-speak, render a massive resolution (better than any monitor) 3D image in real-time, and even perform complicated image enhancement algorithms on the fly.