Bruno Latour, in conversation with Richard Powers, in honor of HAL:
BL: To tell you the truth, I have never understood the Turing Test to begin with. In theory, it should match a flesh and blood human against a silicon machine. In practice however it matches a flesh and blood machine against a flesh and blood machine, so how could the test ever be negative? The distribution is different, I concede. On one side, you have one body explored by ten thousand biologists, cytologists, and neurologists, while on the other side you have one computer concentrating the brain power of ten thousand engineers, software writers, and wafer printers. But how can any Turing Test judge hope to disentangle these two collections? The idea of a test matching a naked, isolated intelligent human against an isolated naked automated machine seems to me as unrealistic as imagining than we are here alone talking through email “naturally”, “directly”, without any mediation. Things and people are too much intertwined to be partitioned before the test begins, especially to capture this most heavily equipped of all faculties: intelligence.