Monday, October 24, 2011

Artificial Intelligence

I was doing some reading about artificial intelligence (AI) in conjunction with my reading of Neuromancer, and I stumbled across some really interesting (although arguably frivolous) creations. The first of which, designed by a Vietnamese robotics firms called TOSY, is a ping pong playing robot. Apparently, the robot has undergone two remodels and is now in its third form, TOPIO 3.0. This robot played against humans at a robotics fair in Tokyo recently, and is said to have held its own. This is absolutely amazing. Ping pong is (and I speak from experience) one of those sports that requires insane amounts of coordination, and TOPIO 3.0 was able to return 10 shots in a row playing against a human. One might say this is a waste of money, or a waste of robotics expertise, but if you ask me, this is astounding.
                                                                            
Another such artificially intelligent creation is the famous Jeopardy-playing robot Watson. As we all know, Watson dominated its competition in each and every game of Jeopardy it played, beating out even the stars Ken Jennings and Brad Rutter by large amounts of money. Again, I find this truly fascinating. Upon doing some research, I found that Watson contains 90 IBM servers and 16 TB ram (for those of us who aren’t computer savvy, that’s about 2000 times the amount found in the average high end NC State laptop (most 2011 era laptops have 4 or 8 GB ram standard). This is ridiculous. Apparently Watson can process the data of roughly a million books in a second.

I think that if you consider how much raw data can be interpreted and used by supercomputers in a short amount of time, it will appear inevitable that at some point, computing technology will reach the point where it can interpret things on its own. I have no idea how or when this will happen, but I am interested to see how long it takes for Neuromancer to become even more accurate. I don’t really think that improving artificial intelligence to unheard of levels is that far off; I think it will be sooner than later. Hopefully our version of Turing keeps a strong hold on it, if it turns out to be selfish or malicious. This is one of those things that is really interesting just because no one knows what exactly will happen.

No comments:

Post a Comment