Well, even with a high level of understanding, any sufficiently advanced use case will still have some uncertainty regarding its "feasibility". Of course, you m

Well, even with a high level of understanding, any sufficiently advanced use case will still have some uncertainty regarding its "feasibility". Of course, you might think that some problems are "solved", e.g., OCR, translation, (common) object recognition, but MANY other problems exist where, no matter how experienced and knowledgeable you are, you can only have an educated guess as to whether a given model can achieve a given performance without actually trying it out. Where experience and knowledge really pays off is in telling apart model performance from bugs. There is a real know-how in troubleshooting ML pipelines and models in general.