Automata in the Looking-Glass

Abstract

In one of his philosophical essays, while broaching the thorny question of ontological commitment implied in natural science, the famous Austrian physicist Ludwig Boltzmann ventures a bizarre, as well as original and unpredictable, description of an ideal device capable of simulating human reasoning and behaviour:

Let us imagine that there can be a machine which appears, behaves and moves like our body. Inside it there is a constituent part which, owing to light, sound, or whatever else, [is able to receive impressions] through organs built exactly like our organs of perception and the nerves linked thereto. This constituent part should also have the capability of retaining the images of these impressions, and, thanks to the intervention of these images, of stimulating the nerve fibres so as to make them produce movements which are identical to our body’s. [...] They will say it is clear a priori that, if, on the one hand, this machine externally behaves like a man, on the other hand, it perceives no sensation. [...] I believe they will say so just because they only think of a clock, rather than of such a complicated machine [...]. In the above-described imaginary machine, every sensation would be simple, each would be identical to complex material phenomenon; only those who do not know the machine’s structure would be unable to measure sensations through their duration and magnitude and would rather represent them through spatial and mechanical images just as we do with our own sensations. However, experience tells us nothing more. Our machine could realize whatever psychic perception experience gives us. Anything else, it seems to me, can only be thought of arbitrarily. Our machine would say, like any man, that it is conscious of each existence (that is, it has mental images because it exists). Nobody can prove it is less conscious of itself than man.2 Of course, consciousness cannot be defined in such a way as to make it concern this machine less than man.3