0.10 Can artificial intelligence be artistic?  (Page 3/3)

If it is an emotional reality then it could be dificiult to figure out, obviously since emotions are subjective, and hard to measure. For instance if you want to relate how cool two different things are it would be impossible because one person could think that it is cool and another person could think it is not cool.

What kinds of emotional reality are measurable and less subjective then? How does emotional reality relate to artificial intelligence? If robots just need to assess objects and how objects are related to each other then they don't need to necessarily have any emotional intellignce.

If a robot needs to understand stuff, it doesn't need to assess emotions, it just needs to understand how to get around in reality. Understanding emotions is too complicated and only a human being could make subjective judgements.

Designing Robots

Just have the robots do stuff - it can't be that difficult to have them function properly. On one hand humans perform very simply -their actions are 'artistic' and can be performed by understanding their own feelings. On the other hand, their actions could be considered to be complex and cognitive and requiring conscouis effort and thought.

If it requires conscoius thought then the person needs to think about what is happening and how to perform in each instance - unless it is already learned (but metacognitive theories already addressed those ideas).

How is a robot supposed to understand that a house is a house, and that houses are where humans live? That seems like something a person would understand, however a robot wouldn't need to understand something like that. What would a robot need to understand that is difficult or complex then, in order for it to be intelligent?

It needs to be capable of abstracting more, robots can be programmed to do lots of things - however that does not make them very intelligent. How could 'abstraction' be programmed into a robot? Abstraction is artistic and uses a humans feelings.

Robots need to be capable of organizing objects in an efficient manner, they don't necessarily need to understand emotions. How could understanding emotions be related to organizing objects, however? Maybe significant objects could generate more emotion, so the significant objects could be related to other significant objects and the insignificant features of an area be disregarded.

The significant features of an area are the features that are, well, significant. They can be significant for different things, different categories such as houses, interiors of houses, a category of objects that consists of things related to human beings, and so on.

What about the definitions of objects? An object could be recognized and its definition looked up by the robot. That way they could form associations - and human minds could be looked at to see what the typical associations are - does this object relate to that object - it has it in its definition after all.

How could a robot relate one object to another object? It looks up the definiton of the object then compares the related features of the two different objects by looking at visual similarities - what about the amount of light shed on objects in a three dimensional space? If the robots can see in three dimensions then they can at least figure out the spacial significance of objects.

If a robot knows the spacial significance of objects then it should theoreticaly be capable of doing anything.

The definition of each object can be programmed in, and how it would relate to other objects and the significance of each object. If the definiton can be described to the robot properly, then the robot should understand everything - however, some things are very hard to understand and it could be dificult to describe it to the robot.

The robot can relate different things to each other, and fit them together spacially so it knows what to do in reality.