What should we regret?
No shoulds, only coulds.
Every action, even inaction, has consequences. Some of those consequences we can reasonably foresee, most we cannot.
It seems to me that what is powerful is to notice what works, in which contexts, and with what reliability.
Use that information to develop ever more reliable models of what can be predicted, and what cannot, and within the class of those things that can be predicted with some sort of reliability, to develop ever more useful models (as time and resources allow).
And it seems that this thing called life comes with various sets of imperatives in respect of time and resources, that act as constraints on the sort of effort that can seem reasonable to us to put into the development of models in different areas.
Sometimes choices seem obvious, other times not at all obvious.
Such seems to be the nature of both the complexity within which we find ourselves and the necessary simplifications we must all make to be able to make any sort of sense of anything in any reasonable timeframe, and even more critically, to take actions required to ensure survival in situations that threaten our survival.
Our brains seem to have evolved many levels of systems that cause us to make very simplistic models in times of threat.
Our news media have discovered that presenting us with information that makes us feel threatened tends to hold our attention and tends to encourage us to buy newspapers.
This is an unfortunate combination of circumstances.
It leads to a high probability of most people staying in the more simple classes of explanatory frameworks available in the class of documented frameworks, and very few people out actively exploring frameworks beyond the boundaries of the documented sets.
This leads to a certain level predictability from the perspective of those interested in political control, but at the same time increases the systemic risk to society as a whole of “Black Swan” events – things not present in the past, and therefore not anticipated and not available to historical analysis (we have an exponentially increasing set of such things).
So, optimal outcome seems to be – No regrets, and take all available opportunities to learn whatever lessons one can from every experience.
And there is an odd trick of the physiological processes of memory, that the higher the emotional context of the circumstance, the more likely we are to recall any particular event – so in this context, and in the sense that regret is an emotional state, there does seem to have been a certain evolutionary utility to us having this emotion of regret – it helps us to remember important things.