Interesting. I wonder if ‘Deep Learning’ is going into the ‘Trough of Disillusionment’.
It appears to be both innovation trigger and trough of disillusionment. This is because humans don’t universally hold the same opinion, and because deep learning will prove to be insanely useful in some areas and deceptively useless in others. It wouldn’t surprise me much at all if the recommendation system itself is a fundamentally intractable problem.