I said that even if it was scientifically proven it still wouldn't be OK to discriminate. Reading comprehension. And the discussion was NOT about systems using

I said that even if it was scientifically proven it still wouldn't be OK to discriminate. Reading comprehension. And the discussion was NOT about systems using correlation as a heuristic for causation. If a company filters users based on something, only correlation matters for their purposes. Believing correlation is only useful as a heuristic for causation is ignorant. For some purposes, correlation matters. For some purposes, the root cause matters. Not everything is the same.