Recently I’ve been reading a lot about AI and ML. It’s quite clear to me what ML is - solving classification and regression problems. There are some fuzzy edge

Recently I’ve been reading a lot about AI and ML. It’s quite clear to me what ML is - solving classification and regression problems. There are some fuzzy edges, but that’s true of any discipline. Maybe you invoke Mitchell’s definition and say “well, it’s improving on a task” (as many introductions to ML do) but that’s completely out of step with what people actually treat as ML. There are lots of interesting “learning automation” areas that we’re neglecting - symbolic reasoning being the glaring one for me. AI just seems like a nonsense term to me. Maybe it’d be better if we stopped using it.