Probabilistic Thinking, Probabilistic Reasoning
Fermi paradox is similar in that since there is only one case, we can only estimate the probability based on the conditions of that case.
Therefore, we must fundamentally view the occurrence and non-occurrence of events as a spectrum.
Errors in statistical estimates about the past occur because they are made using current data. If this current data strongly depends on past events, it can lead to overestimating the probability of that specific past. In other words, since we only have information about the established past, it's difficult to rationally consider other possible paths. Therefore, we tend to assign higher probability to the time path that led to our current data, which is the only information we can reference. This error is more likely to occur when the amount of information is insufficient compared to the time interval being estimated.
Bayesian, Update yourself incrementally
- Correlation does not imply causation
- Uncorrelated does not imply independent
I prefer using statistically generalized expressions directly. Otherwise, I have seen many cases where propositions that hold statistically but not in specific situations, such as "men are stronger than women," either receive unnecessary additional explanations or focus on specific examples that mislead the logic. - Seonglae Cho
