Anti-Proof of Artificial Intelligence & Accident

The goal of Strong Artificial Intelligence is to make a machine which while composed from simple unintelligent components, replicates the intelligence of human beings.

Intelligence is the quality of humans which scientists and researchers would use to create such an artificial system.

It therefore follows that a truely intelligent system would have the qualities required to discover and create itself, just as the human creators have done.

And if it has the qualities to understand and create itself, then it also has the capabilities to understand and create a human being intelligence.

We know that humans were not created by an intelligence but rather by natural unintelligent forces.

If the Artificial Intelligence has the capacity to create itself, and thereby the capacity to understand and create humans, then it also has the capacity to understand the unintelligent forces that created humans. It has to have an understanding of Natural forces also, not just of logic and intelligence subjects.

The loop is infinite. True Strong Artificial Intelligence requires a machine so vast that it can understand the forces that created the universe, and then its own creators and then itself. Such an Artificial System is neither artificial, nor a system - it is the universe.

If on the other hand we decide to take the path of intelligent design, our Artificial Intelligence must replicate the intelligence of God, it must become God. And God himself must have the intelligence thereby to understand and create himself, ad infinitum also.

This is a demonstration of a principal I wish to call "the inefficiency of self-reflexivity".

If you photograph a camera in a mirror, is the picture taken really the camera? Yes, but it is not identical because now you have a picture and the real camera. There is no system where the picture taken is identical with the picture taker, because if they were which came first - the picture taker or the picture? If the exist simultaneously where is the difference - aren't they just the same thing?

In computing this is the difference between "data" and "instruction code".

Another analogy is the proof that there is no best algorithm for compressing data. The minimum compression is the same as extraction the information from a system. Many systems have redundancy created by repetitions. This physical sentence has many repeating words and letters and can be enormously compressed to extract the information. So there is no algorithum to extract all the information from a system either. The proof is basically like this and is another example of the inefficiency of self-reflexivity:

Were such an algorithm to exist, it could be coded into a computer and then used to extract it's own information content. Such an program code would have to take an input, namely the code to be compressed. So we would have to code the algorithm with variables, and then to use it replace the variables with the code. But that code would need it's variables replaced and so on infinbitely. This is not the actual rigorous proof, but it illustrates the point.

Searching for a completely closure on understanding is an endless task. Research will only ever succeed by accident in discovering Artificial intelligence, in its simplest way by replicating the biological processes in childbirth.

Accident indeed has been the foundation of human progress from (apocryphally) Newton's apple to Alexander Fleming, and is the magical force which gives us the world around us. "Accident" cannot be owned, predicted, or understood - indeed it is the magical principal which blows apart the myth of Intelligence and Artificial Intelligence - the nearest actual thing to God.