At the moment, robots can’t think on their own – they need to be told what to do by a program and that program is made by a human. The program will tell them what decisions they need to make in certain circumstances (e.g. if they are travelling forward and there is a wall there, they will need to turn around or reverse) but the robot won’t be able to decide that without being told beforehand.
BUT
Robotics is moving very fast and soon a lot of the decisions that a robot makes will be decided by the robot themselves as they begin to take signals from the outside world (this is much like us when we ‘see’ something we can either pick it up, avoid, wave etc.) It’s very exciting stuff….but a little scary too, don’t you think?!
Not yet, but they aren’t that far off. There is a particular test called the Turing test which you can use to see if a robot/artificial intelligence can “think” for itself. None has passed yet, but there are some that get quite close.
One of the areas that I am interested in, though have never worked in, is the area of neural networks. This is where a computer is built like a brain, imitating how it works. Using this system you can train the computer with some information, and it learns how to understand that information. Then you can use the computer to search a lot of similar information and it will be able to process it a lot faster than any human. It still can’t think for itself, but it is able to find answers much much faster.
Comments