I was talking to a professor who made a comment to my question.
Me: So much of quality literature around this topic ( IP Protection for Neural Weights) emanated in 1990-1991, I'm truly at loss to understand what went wrong. Why did everyone suddenly stop worrying about this topic?
Prof: I think AI was starved of training data and what training data there was demonstrated that, depending on the architecture of the Neural Network ("NN"), there would be some NN's that could not be trained. They would essentially oscillate and never reach equilibrium. That is where the "Deep Learning" approach solved the problem -- and allowed AI/NN to move forward.
Fascinating. I looked up and indeed Deep Learning pushed forward in 2000 after it was introduced for Artificial Neural Networks by Igor Aizenberg and colleagues in 2000.
I want to collect research papers/ opinions around this point. Was AI/NN really bottlenecked until 2000 (Advent of Deep Learning)? What spurred the exponential growth in this domain?