Re: learning with constraints


JGCASEY wrote:
> Unfortunately I cannot translate behaviorese into
> a program but people have translated 'careless
> talk' into programs. Glen can use his behavioral
> talk to declare that the Roomba robot's behavior
> is mediated by its electronic brain while I shall
> talk about *how* the Roomba robot "decides" what
> to do next using good old careless talk we all
> can understand down to the last logic gate.


To be fair, you have to admit that so far the entire field is guided by
dumb evolution, natural selection of "best of worst" ideas. Things you
talk about in "carelesse" are lucky guesses, at most. There's simply no
solid theory that would allow for artificial selection in the field.
Yes, we have nuclear weapons now, but few centuries back atoms were
just something they have talked about in "philosophese".

.

Relevant Pages

  • Re: learning with constraints
    ... >> Unfortunately I cannot translate behaviorese into ... >> a program but people have translated 'careless ... >> is mediated by its electronic brain while I shall ... > allow for artificial selection in the field. ...
    (comp.ai.philosophy)
  • Re: learning with constraints
    ... >> without constraint would make every character ... behaviorese as everything else is 'careless talk'. ...
    (comp.ai.philosophy)