I’ve actually been recently thinking about randomness in ML, and I’ve come to a compelling case for it’s specific role. The insights do seem to generalize to all problem-solving mechanisms in a way. I can expand if you want
Sure
Sure