Given the remarkable brute force power we have access to, namely lots of data and computational power, is causal inference the “light weight and feather” of cognition?
All this talk of causality is needlessly confusing. The book of why is 5% useful ideas, 20% bad explanation, and 75% about how amazing Pearl is and how dumb everyone else has been.
Causes have effects and not vice versa.
But correlations are just that: lines.
Causation does not imply correlation; it implies association, and associations are asymmetric. When relationships between data are asymmetric, proper causality is already baked into the system.
All this talk of causality is needlessly confusing. The book of why is 5% useful ideas, 20% bad explanation, and 75% about how amazing Pearl is and how dumb everyone else has been.
Causes have effects and not vice versa.
But correlations are just that: lines.
Causation does not imply correlation; it implies association, and associations are asymmetric. When relationships between data are asymmetric, proper causality is already baked into the system.