# Feedback loops and play are important for breakthroughs However, it needs to be in a [[Goldilocks]] situation. Too much play and you’ll never do anything. Too tight feedback and you’ll end up in an inadequate equilibrium. If you think about it in terms of machine learning or optimization, the tightness of your feedback loops are like your learning rate and play can act a bit like simulated annealing. ### Related * [[alexanderStudiesSlack2020]] <!-- #stub --> [Web URL for this note](http://notes.benjaminreinhardt.com/Feedback+loops+and+play+are+important+for+breakthroughs) [Comment on this note](http://via.hypothes.is/http://notes.benjaminreinhardt.com/Feedback+loops+and+play+are+important+for+breakthroughs)