why don't we ban complaining about the semantics of "AI" too. Just about every thread on AI has someone saying that it's just layered regression or some similar rank dismissal. It's like saying calculus is just applied limits, true enough but not enlightening or useful.
imo I'm fine with any of these being called "AI"
- neural nets, deep or otherwise
- any classical ml techniques like decision trees, svm etc.
- any classical nlp techniques or tools
- "modern" nlp, eg. glove, fasttext etc