Artificial intelligence seems like a modelling program for interfaces of possible processes forming into another possible or greater architectures of further or differing processes. In what capacity is it intelligence and not simply a continuous flux of ever-increasing builds of emergent information that ISN'T necessarily feasibility in function but rather feasibility of new emergent abstractions? More clearly,how does artificial intelligence create valid data we can deem as intelligence as oppose to simply new avenues of abstraction that are fixed to our criteria of validity?
The term AI, short for Artificial Intelligence, is misleading because it hides the technical details. In the mainstream assumption, AI is some kind of intelligent software which acts autonomously and replaces a human operator. But that is not the principle how AI is created in software. A working AI system is mostly a human-to-computer interface and not an autonomous software. It's inner working is similar to what a computer mouse can do. Nobody would argue, that the mouse pointer on the screen has some kind of intelligence, but in reality the mousepointer acts on a human-level.
The philosophical debate around Artificial Intelligence is driven by competitions in which the creation of an AI system is desired. The game consists of many instances. At first, the moderator is starting the AI challenge, he asks for an AI. Then the participants are creating software. Their AI software works internally as a human-machine-interface. The feature of the software are evaluated by the moderator of the competition. What I want to express is, that the term AI is connected to a competition in which the creation of an AI is mandatory. The definition of what an AI is depends on the competition in which the AI has to be created.