---

Turing's original imitation game paper, which I'd never read until the other week, gets very strange indeed towards the end, but this may just be a product of my post-corporal-punishment schooling.

He says, after proposing a genetic algorithm (is this the first appearance in the literature?), of 'unemotional' methods of communication:

If these are available it is possible to teach a machine by punishments and rewards to obey orders given in some language, e.g., a symbolic language. These orders are to be transmitted through the "unemotional" channels. The use of this language will diminish greatly the number of punishments and rewards required.

There is a well-established field of machine learning. What could machine punishment look like?

posted at: 14:20 | path: /imitationgame | permanent link to this entry