What we commonly refer to as “Artificial Intelligence” is not true intelligence at all; rather, I see it as an encyclopedia equipped with a calculator.
The distinction between an encyclopedia with calculating functions and a knowledgeable human lies in how they connect with the world. A knowledgeable person engages with the world as a human being, but does an encyclopedia with calculating functions connect to the world as such? Can this ever transcend the boundaries of human programming?
When we view AI as a calculating encyclopedia, both its spontaneous generation and its hallucinations can be seen as errors in computational results. This is because the criteria for “legitimate generalization” are internal standards within the program, rather than external ones.
