Earlier tonight, IBM’s supercomputer “Watson” went head to head against Jeopardy’s two top players in history, Ken Jennings and Brad Rutter. The system that has been in the making for 3 years and was specifically designed for the Jeopardy competition, tied with Brad at $5000 at the end of the 1st night’s competition.
After nearly a decade, Oracle’s copyright lawsuit against Google is close to settling an important question: can you own the basic commands of a coding language
IBM has a patent out there for a smartwatch packed with a bunch of folding screens, because the future is dumb as heck.
DCD reports on the epic decades-long quest to make computers more like the human brain. Early efforts brought us the deep learning revolution, neuromorphic computing could bring us so much more
It's so interesting to see AI becoming to competitive, but I wonder if it will become self aware one day once the code becomes complex enough, as in being able to learn and add to its code to better itself.
Fun to watch - kinda eerie though to clap for a supercomputer :)
the start of sky net
There is no such thing as an empty learning algorithm.
There always needs to be source data to compare too, in order for the algorithem to weight choices. Neural Networks (programming)show this as an example.
Think about putting a child in a box or in a room and locking him in. Then tell him, get out of the room by any means neccessary, however you deem fit. You could leave several instruments in the room or hide a key etc... A computer will never know all the variables in the choices unless it is able to (On its own) identify a crowbar as a choice way of excape, an apple as NOT a choice way of escape, a Ladder as a choice and a dog as NOT a choice etc...
But even further then that the child could just percieve that crying would be the best way out. So they cry and sit in a corner and when the person comes in to check ot hem, they quickly dash out.
All of these things are unaccounted variables and for a computer would have to be preprogrammed.