The driving force behind generative models

From distributions to point estimates and back

And why Bayes’ theorem is a neat way to work with probabilities

What does it mean to have a trustworthy model?