3 ms·
> There are testable positive claims - things like the mirror test. Yes, but the problem with the mirror test (and all similar tests) is not the observation bu
by lutusp 14y ago
> There are testable positive claims - things like the mirror test.
Yes, but the problem with the mirror test (and all similar tests) is not the observation but the conclusion -- one in principle could program a computer to recognize itself in a mirror, but have no other properties we assume define "consciousness".
On could in principle create a list of properties of consciousness, then have a devil's advocate program a robot to display all the required properties -- just to prove a point.
In DARPA road races, robot cars manage to avoid each other and fixed obstacles, mimicking the behavior of human drivers. It's only a matter of time before a car will pass a sort of "Turing test" and be indistinguishable from a (sober) human driver, i.e. be "conscious".
> dealing with things that are hard to define doesn't mean something 'is not science'.
In fact, yes, that is a legitimate argument for calling something non-science. If independent observers cannot agree on their terms, if similarly equipped observers cannot draw the same conclusion from a given observation, this prevents the consensus on which science depends. Science is, after all, ultimately a shared, falsifiable theory about the world, therefore deconstructive post-modernism (the idea that all observations are subjective) cancels science.
> People didn't run around saying 'oh, the atom is just philosophy' ...
They had the right to do that until there were unambiguous observations, and testable theories about the observations, on which different observers could agree. Until then, atoms were philosophy.
Aristotle once claimed that women were inferior to men because they had fewer teeth -- hard to believe but true. It remained philosophy until someone started counting teeth.