We absolutely need low-stakes mechanisms to discover potential causalities to research further.

What we need to stop is put unreasonable emphasis on association studies (i.e., interpret the results incorrectly).