Sitting on a boat in Clear Lake the morning of my now-sister-in-law's wedding, somehow I felt compelled to write this mixed metaphor about optimization and the pursuit of truth:
The canonical analog for truth-finding is that it’s like multi-variable
optimization. You seek the highest point by heading uphill and search until
there’s nowhere higher to climb. I think this analogy is flawed. It’s flawed
because we know which way is up, at least, if we’re to have any hope that we
can be successful, we have to assume that. It should be easy, then, to compare
the levels of different local maxima to determine which is the global max. Yet
it’s not. The problem seems to be that we don’t know which way the truth axis
should be oriented, so we attach to different maxima which seem more right with
our personal orientation. So how do you determine which orientation is correct?
I’m not quite sure. Some truths seem to jump out and assert themselves loudly
and clearly, while others are highly counterintuitive and may seem downright
wrong at first. It would be good to have a way to self-determine the difference
between counterintuitive good ideas and bad ideas without resorting to just
doing what other people say. Listening to others is okay, just do it
thoughtfully. Must investigate further.
No comments:
Post a Comment