Is there any theoretical AGI modell based on Popper theories? I've heard only about ones based on Solomonoff induction & bayesian inference, like: https://en.wikipedia.org/wiki/AIXI and its various derivatives...
Re:Progress in Philosophy
by abecedarius on Sunday October 07, 2018 @10:18AM
(#18160642)
There was an old paper "The scientific community metaphor" which I thought was pretty interesting and Popper-flavored, though it was more of a sketch of a GOFAI architecture than anything like AIXI.
Yes; if you like a philosophical idea that's usually sufficient for you to have 'adopted' or 'accepted' it. Yet, in addition, philosophical problems have been solved. It's just that ipso facto they then become part of science or some other discipline, so we don't notice so easily. For example, the theory of evolution was an example of philosophical progress (a new way of thinking about something, in this case life). But it's now part of biology. Another example is that Karl Popper solved the problem of induction. It's controversial that he did so. But if it shelves off into some other field, e.g. AGI, then perhaps his achievement will be more widely accepted as such.