Prediction: Any AI problem that you can simulate and sample endlessly many training samples for can be solved with today's algorithms such as deep and reinforcement learning.
— Richard (@RichardSocher) August 6, 2018
Prediction: Any AI problem that you can simulate and sample endlessly many training samples for can be solved with today's algorithms such as deep and reinforcement learning.
— Richard (@RichardSocher) August 6, 2018
This is meaningless for problems where the required number of samples grows exponentially. Remember that Go and Chess are played in a fixed-size board.
— Thomas G. Dietterich (@tdietterich) August 6, 2018
"this" meaning sampling and simulating unlimited numbers of steps. My experience applying RL to wildfire management suggests that the number of required simulations is infeasibly large even for the massive farms Google is using, hence the claim is meaningless
— Thomas G. Dietterich (@tdietterich) August 7, 2018
Just because it doesn't apply to every problem doesn't make it meaningless. It also doesn't apply to language understanding because we cannot simulate language. Which is why techniques that have this requirement probably won't get us there.
— Richard (@RichardSocher) August 7, 2018
+1. Yesterday, I had a painful conv with some exec of super well-funded startup. He seemed so strongly convinced that dialog (chatbots for eg) can be “solved” using reinforcement learning. Any of my attempts to disabuse him of that notion made him more entrenched in that position https://t.co/1CIhtXbpRd
— Delip Rao (@deliprao) August 7, 2018
Prediction 2: The currently hip ai algorithms that require endless samples from a simulation will never get us to generalizable ai capabilities (or agi).
— Richard (@RichardSocher) August 9, 2018
Some things can't be simulated/sampled until solved like natural language or many other important areas like medicine.