[ 3/November/21 ]
There are a lot of things to consider when considering this question.
How is it that we think?
What are the multiple levels of the roles of various sorts of biases in the ability of our neural networks to make any sense at all of the reality in which we seem to be embedded?
How deeply have the evolutionary pressures to supply reasonably reliable solutions to common existential problems biased us to prefer simplistic certainty over more complex levels of confidence, given that many contexts require rapid solutions with as little energy as possible, and the slower tend to be punished much more harshly than the slightly inaccurate?
How deeply have the many layers of such pressures in our sensory and decision making and valence systems led us to experience reality as something vastly more simple than it actually is?
Does simple binary logic (True/False) really represent how reality works, or is it just that it is the simplest of all possible logical systems, and therefore the one that evolution happened upon first, and biased us to prefer?
Can reality be more accurately modelled in the next simplest of logics (trinary – True/False/Undecided)?
Could it be that reality is actually deeply more complex that either of these simplest of possible logics allows for?
What does one actually mean by randomness and chaos?
If one takes the mathematics of quantum mechanics at face value, then it tends to indicate that all things are related (in a probabilistic sense), and that in some contexts some things are definitely far more probable than others, and all contain fundamental uncertainties.
There is a logical necessity, for complexity such as we are to be able to evolve, that at least at some scales, and in some contexts, that the world does very closely approximate classical causality. Complexity such as we are cannot exist without reasonably reliable sets of constraints over time. But such reliability need not be 100%, and need not extend across all scales or contexts.
It seems to be the case that only in a universe that contains some reasonable approximation to classical causality can organisms such as we are evolve, but only in a universe that also contains some degrees of fundamental unpredictability in some contexts and scales can it also be true that complex organisms such as we are can have real, meaningful degrees of freedom and choice. And such freedom cannot be absolute, because freedom from all constraints necessarily destroys the levels of complexity that made such freedom possible. Freedom must always have limits (responsibility) if it is to survive (all levels, always).
So it seems clear to me, as someone who has been examining various sets of evidence, and various ways of assembling and considering evidence, and various sets of hypotheses for well over 50 years, that it seems very likely that this universe does contain fundamental unknowns and uncertainties, at multiple levels, and can also very closely approximate classical causality at some levels, and thus we can have both engineering and morality – and both can be necessary and reliable to reasonable degrees.
It also seems that whatever objective reality actually is, it is more complex than any computational entity could possibly accurately model in anything remotely approximating real time, and therefore all models or understandings of it are necessarily wrong, and it is the tools and techniques of science that allow us to most reliably become less wrong over time. So the appropriate way to approach any set of “knowledge” is with a little humility and caution and questioning (as contexts allow).
Even the simplest of models can be close to an optimal solution to the problem of “survival in uncertainty” in some sets of contexts, and wildly inappropriate in others – same goes for the more complex ones.
So it seems logically inescapable that ignorance will be our eternal companion, should we manage to live for the rest of eternity. And that leads to the seemingly paradoxical situation that the more certain someone is that they are right, the less likely it is that theirs is actually a reasonable approximation to whatever reality actually is.
To get a tiny appreciation for how confidence can come out of uncertainty, consider time. The smallest time a human can perceive is roughly about 1/100th of a second. We can build atomic clocks that seem to us amazingly accurate, and the cesium atom cycles about 10 billion times a second, that is about 100 million times in the shortest time a human can experience. So to us, that seems a very short time, and atomic clocks seem very accurate. But on the Planck scale of quantum mechanics the time unit is about 10^-43 seconds. So that means that there are more Planck time units in a single cycle of the cesium atom in a cesium atomic clock, than have existed cycles of that cesium atom in the age of the universe. So any uncertainties in that Planck scale time become a very well populated distributions that behave very close to classical causality even at the scale of an atomic clock. And there can be contexts where such uncertainties can have real world impacts at our scale, and that is a deeply complex topic.
Another important consideration is that, contrary to popular dogma, it seems that the evolution of all levels of complexity is founded upon new levels of cooperation, and thus for entities such as ourselves, with all the multiple levels of complexity embodied within and around us, it is much more accurate to characterize evolution as being mostly about the emergence and stabilization of new levels of cooperation. And raw cooperation is always vulnerable to exploitation, and thus requires emerging ecosystems of cheat detection and mitigation systems. Recurs that to every level you are able.
So it seems to me that randomness and chaos are real in some contexts and to some degree, as is ignorance, as is causality – and all contain contextually relevant matters of degree (necessarily).