How do you quantify that? Is it the downdoots? Or is it the degree to which I'm willing to waste my time discussing shallow inputs? Or do you go by gut feeling?

On a serious note: Have you heard of brandolinis law? The asymmetry described there has been shifted by several magnitudes with generative "ai". Unless we are going to start to use the same models to argue with people (e.g. chatbot output) on the net, we will have to choose much more carefully what discussions we involve ourselves on, don't you think?