How does a robot think?

Answer 1

"They do not think. They can be programmed to react to changes in variables only."

Going by that logic, humans don't think either ;). If you have something burning your hand, you can imagine a specific variable, called god knows what, but we'll call it MyHandIsBurning, and it can be true or false. If its true, then you will always yank your hand away from wherever it was when it became true. That's reacting to a change in a variable. Although it is infinitely more abstract and complex than just that.

A queen bee in a bee's nest maintains the entire order of the colony with 2 simple rules, 1 - find empty places to lay eggs in, 2 - stay near the center of the hive. For years it appeared that a complex thinking process must be occurring in the queens brain to keep track of the structure of the hive, when in actually it was a simple rule based behavior, incredibly robotic.

A robot doesn't think, but goes about its life reacting to things, with no forethought into what may occur afterwards. We think because we take into account how an action may affect us in the future. When driving toward a brick wall, you may be able to turn right or left, but you wouldn't want to react in a way that would lead to a situation where, after turning to miss the wall, you drive off a cliff.

If a robot goes up to that wall, and observes that on one side there's a cliff, and the other side a safe stretch of road. Providing it was programmed to detect cliffs and know they're bad, it will turn the right way. To anyone who didn't know exactly how this was working, it would appear to be thinking.