2 ms·
If I have no way to evaluate what is objective, than how do I know that I'm closer to objectivity? At the point that we conclude "there isn't a single objective
by Der_Einzige 7y ago
If I have no way to evaluate what is objective, than how do I know that I'm closer to objectivity? At the point that we conclude "there isn't a single objective thing in the whole universe" we have no reason to believe that the appearance of getting closer to objectivity actually correlates with getting closer to objectivity. It's entirely possible that "objective" or a global optimum solution is extremely "close" to a solution which appears to be extremely suboptimal. Science finding what appears to be good local solutions using rationality is the same as finding "lottery tickets" in a state space for a problem.
This is why global optimization methods like Simulated Annealing or Genetic Algorithms rely so heavily in efficient exploration of a state space, and why they can easily solve adversarial reinforcement learning problems that backprop techniques (get "global solution" by moving closer to locally better solutions) fail at (the cart hill climbing problem being the classic one)
I'm willing to believe that there are local "solutions" which appear to get better solutions to a problem than random, but I have no reason to believe that my solutions which appear better correspond to actually better solutions (I may have failed at evaluating the correct measure of "good") , or even that by moving towards the appearance of a better solution that it gets me closer to the global solution (which is why backprop fails on cart hill climbing)
tl;dr the post modernists are right on this one and science needs to immediately dispel it's platonic rationalist roots because it's basically just applying backprop (Plato would call it the "dialectic") to a philosophical problem and hoping that human thoughts gets us closer to truth because it appears rational to us. Yeah, actually, you should put the astrik on all science that says "may not actually correspond to truth because we don't know what truth is"