3 ms·
To be clear, I didn't downvote, but I guess people find it hard to believe that you sit down to do the calculations for a Bayesian update of your belief state f
by _Algernon_ 3y ago
To be clear, I didn't downvote, but I guess people find it hard to believe that you sit down to do the calculations for a Bayesian update of your belief state for each piece of information you encounter.
I assume that your "statistical approach" translates to a much more haphazardly approach that's similar to how other people do it (gut feeling, based on rules of thumb and heuristics), and describing it as statistical is disingenious.
- c22 3y agoI guess that's fair. I don't really run any numbers when I update my beliefs because usually the "winning" hypothesis enjoys near-saturation of my trusted sources and often the "plausible but novel rationalization" turns out to be a unique snowflake in the landscape. If I notice the same novel argument getting picked up from somewhat-independent multiple directions I revisit it and apply a more rigorous rethink. This is what I mean by a "statistical" approach. Perhaps some other word would have been better. I do, however, pretty regularly sit down with some graph paper or an envelope and review/diversify my trusted sources.