Occam's razor isn't empirical, but it is the economically rational decision when you need to use one of several alternative theories (that are exactly "compatible with the evidence"). Besides, "further observations" are inevitable if any of your theories are actually going to be used (i.e. to make predictions [that are going to be subsequently 'tested']).
This thread is for the discussion of Less Wrong topics that have not appeared in recent posts. If a discussion gets unwieldy, celebrate by turning it into a top-level post.