Comment by danielbln

Right now devs around the world are pushing ungodly amounts of reinforcement learning data into the big AI labs. There is no reason to believe these models won't handle this stuff themselves and our priors will become a useless relic.