13 ms·
I don't think rationality per se is overvalued but our capacity to achieve it consciously may be overestimated while the capacity of our brains to unconsciously
by tgflynn 7y ago
I don't think rationality per se is overvalued but our capacity to achieve it consciously may be overestimated while the capacity of our brains to unconsciously generate reasonably accurate approximations to rational decisions may be underestimated.
It's clear that there is far more information processing occurring in our brains than we are ever consciously aware of. We end up being conscious only of the final result and we tend to call this "intuition". It would be a mistake to dismiss using intuition as irrational just because we aren't conscious of the mental activities that generated it. Of course intuition isn't necessarily right, it needs to be consciously evaluated and we should also use our conscious minds to train our intuition to increase its accuracy.
- philipov 7y agoWe don't trust intuition for the same reason we don't trust the output of neural networks, because the process that generated it is inscrutable. It is ironic that neural networks, an essentially irrational tool, have become so popular in a culture obsessed with rationality.
- tgflynn 7y agoYes, that analogy occurred to me as well. But I don't think a blanket policy of distrust is optimal for either of those "technologies". Both are useful in the right circumstances and trustworthy alternatives are often not available.