Turing's 1950 paper asks, "Can machines think?"
After introducing the Turing Test as a possible way to answer the question (in, he expects, the positive), he presents nine possible objections, and explains why he thinks each either doesn't apply or can be worked around. These objections deal with such topics as souls, Gödel's theorem, consciousness, and so on. Psychic powers are the last of these possible objections: if an interrogator can read the mind of a human, they can identify a human; if they can psychokinetically control the output of a computer, they can manipulate it.
From the context, it does seem that Turing gives some credence to the existence of psychic powers. This doesn't seem all that surprising for a British government mathematician in 1950. This was the era after the Rhines' apparently positive telepathy research — and well before major organized debunking of parapsychology as a pseudoscience (which started in the '70s with Randi and CSICOP). Governments including the US, UK, and USSR were putting actual money into ESP research.
Yes, but also remember that Turing's English, shy, and from King's College, home of a certain archness and dry wit. I think he's taking the piss, but the very ambiguity of it was why it appealed as a rationality quote. He's facing the evidence squarely, declaring his biases, taking the objection seriously, and yet there's still a profound feeling that he's defying the data. Or maybe not. Maybe I just read it that way because I don't buy telepathy.
Another month has passed and here is a new rationality quotes thread. The usual rules are:
And one new rule: