Buried inside @nickcammarata's very cool and encouraging article about AI interpretability, he references another brief article called "the bitter lesson": incompleteideas.net/IncIdeas… In it, Rich Sutton observes that attempts to develop AI capabilities by cleverly embedding pre-existing human knowledge work okay in the short term, but in the long term lose out to generalized computation through search and learning. In other words, optimizing for discovery always beats recombining the ever-repeating known. I can't help but apply this to the human psyche. How often do we wrap ourselves around the axle of optimizing for the "right" way to do things, learn "rules for living", or filter ourselves through layer upon layer of identity, when we'd fare much better by following a more general approach of maximizing search and learning? I can't think of a better way to describe the effect of inner work than that it frees up mental, emotional, and spiritual resources for uninhibited, generalized search and learning. What if doing so also makes *us* more "interpretable", aka transparent and trustworthy, to our own neighboring neurons in our social network? Maybe the human neural network is also, as Nick says, "already clean", and well on its way toward convergence with the structure of the universe.