And like any other learning algorithm free-play has "real-world" information, free exploration, creativity. The vast majority of which is very slowly learning w

And like any other learning algorithm free-play has "real-world" information, free exploration, creativity. The vast majority of which is very slowly learning what not to do. Most times, that's all that happens. Cultivation, when perfectly executed, provides about 50-50 what to do and not to do. Cultivation results in far faster progress. In deep learning, in reinforcement leearning, and in humans. (reinforcement algorithms are worse actually, they will simply barely make any progress at all until they see a positive example) The problem is that this is really, really not what people want to hear. Free play, freedom to explore, ... results supposedly in happier children, more fulfilling lives, creativity, ... Except ... no it doesn't. It's an empty meaningless long string of failures slowly sapping strength until you fail, with very rare example of great success. It's spending seed corn on a lottery ticket. This is made more difficult by the observation that, yes, eventually, if you do climb to the very top of society you'll be playing the exploration game, without any cultivation, regardless of whether you got there with exploration or cultivation.