4 ms·pron makes no claim that neural nets are on the extreme end of ad hoc.by argonaut 10y agopron makes no claim that neural nets are on the extreme end of ad hoc.zodiac 10y ago> like somebody tried lots of things and one seemed to workargonaut 10y agoWhich is a perfectly accurate description of many architectural improvements in deep learning (e.g. residual networks).zodiac 10y agoYeah, but it sounds like the extreme end of ad-hoc.