ABSTRACT The Turing test is usually seen as an operationalisation of the question whether machines can think. In this paper, my aim is to show that by understanding the test in this way, one ends up in scepticism about the existence of minds in general. By focusing on whether some particular machine can pass the test, or whether it can be said to be an accurate criterion for thinking, most scholars accept the underlying assumption of the scenario, namely that we need an empirical procedure in order to secure the knowledge that someone thinks. I argue that the Turing test can not only be fruitfully paralleled to Descartes' language test, of which Turing knew, but should also be related to the former's sceptical method in the Meditations . Such a comparison makes clear why the scenario Turing imagined implies a scepticism not only about others′ minds but also one's own, as it abandons the idea that one can non‐empirically know in the first person that one thinks. I will end by sketching a solution to the sceptical problem of the knowledge of other minds by introducing the concept of second‐personal knowledge that someone is thinking by thinking the very same thought the other thinker thinks.
Marvin Tritschler (Sun,) studied this question.