If Venice reinforces one lesson, which I hope I may pass on, it is this: blindly following logic, the numbers, efficiency, algorithms...
Reductionism is frequently shown to have jumped to false conclusions with each new discovery made.
Building a city doomed to depths, even before global warning, is “the insanity of genius” logic would have advised against, robbing the world of a cultural catalyst for enquiry. I find, often - perhaps more often than not, what turns out to be right at first appears insane.
As Thomas Kuhn enforced: we should not follow the Bayesian logic of the white swans we can see (following the ‘facts’ we have access to); until the world is perfectly encoded, which I hope never happens, the only certain thing is that we must adopt an attitude of uncertainty - the assumption that crucial information is missing.
Only with this attitude can we discover the black swans - primed for a paradigm shift, else confirmation bias rules.
Machine learning techniques are ipso facto Bayesian therefore don’t capture ‘truth’ - waiting for just one outlier to prove them incompatible with reality; they summarise truth.
Caution: abbreviation is the mechanic of unintended consequences and friction to true discovery.