Artificial intelligence means that a computer actually becomes intelligent.
The Turing Test
The state-of-the-art test to detect such intelligent computer is the Turing test. In this test, the presumably intelligent machine and another human are having a conversation with a human tester. If the tester cannot distinguish between the machine and the other human, the machine is assumed to be intelligent.
This was of testing intelligence has major drawbacks. First of all, it is a subjective decision if the human tester determines the state of intelligence. With a reasonably stupid human tester, even a dumb machine might pass the intelligence test. For example, this was subject to critics for the computer program Eliza. Quite a dump computer program, still many humans considered it as human substitute.
A Novel Test
Lately I read the book "Quantenmechanik" by John von Neumann, the inventor of the modern computer architecture, by incidence. One of the puzzling, and as it turns out, still unresolved mysteries about quantum mechanics is the following: for a quantum mechanic system, it makes a big difference if someone is observing it or not.
Even though, this is not explicitly stated, this person watching it is assumed to be intelligent. Especially, the treatment of thermo-dynamic systems clearly shows that the significance of the observer gaining information from the system. It requires conscience to irreversibly collapse the QM wave function! Otherwise, the system is reversible.
Have a test system observe a quantum mechanical system. If the wave function collapses, the system must be intelligent.
Advantage: objectivity, will not be affected by stupid human testers.
What else does this test imply? Can we learn more about the nature of intelligence or conscience by adopting this point of view?