Can psychiatrists distinguish a computer simulation of paranoia from the real thing? The limitations of Turing-like test as measures of the adequacy of simulations.