Quote:one of the primary usefulnesses of AI is that algorithms can be really good at doing things that humans are really bad at.
Fair enough, I don't deny their effectiveness either. But do you really think that goal has little places in AI development ?. Questions have been answered, algorithms have been developed by human only. Wouldn't it even better/be the right time to let
computers raise and solve them, or support us discover things which aren't accounted ?. People can not be duplicated, but computers can be.
Time, money and effort have been put into imitating parts of human intelligent, some have evolved fairly good indeed. Don't they really need these advanced/complex/imperfect algorithms, however ?, or just something really simple that we couldn't find out ?.
I'd like to note that man in general (not to mention disability or unsuitable environments) can be taught to do everything that others can. Why do we have to transform every single algorithms to computers, instead of building a real learning algorithms ?. We could't find it yet, I think.
Sorry to be (if any) off topic.
--> The great thing about Object Oriented code is that it can make small, simple problems look like large, complex ones <--