The eventual completeness of such a model depends on his updated version of the test developed by Alan Turing to identify whether computers could reason well enough that a human communicating with them could not tell whether a human or a machine was at the other end of the communication.