a classical anthropic Everett model

"..the only kinds of entities we know for sure to be
real are our mental feelings and perceptions (including dreams). The material
world in which we have the impression of living is essentially just a theoretical
construct to account for our perceptions."

This was not written by Ernst Mach but is from a remarkable paper by Brandon Carter, dealing with the classical limit of the Everett many worlds interpretation.

"Assuming ... that mental processes have an essentially
classical rather than quantum nature, this essay has the relatively modest
purpose of attempting to sketch the outlines of a simpler, more easily accessible, classical unification ... as an approximation to a more fundamental quantum unification that remains elusive."

As every other many worlds interpretation, it struggles with the issue of probabilities and proposes a solution "based on the use of an appropriate anthropic principle in conjunction with the Everett approach...".

I really like the way the problem is posed, but I don't find the answer very convincing [x]. If one acknowledges our perceptions as being necessarily
classical realities, does Bohr's interpretation of quantum theory (or something similar) not make much more sense?

You, as a sentient being with q=1 (or as just another one of my theoretical constructions), be the judge of that...

[x] One problem I have with it (p.6ff) is this: Once we assume that the outside world is only my theoretical construct, how can the number of sentient beings (which are only theoretical constructs as well) make a difference to the calculated probabilities? Also, would the anthropic argument not indicate that it
is always more likely to survive a certain situation than to be killed?

really the best of all possible worlds?

Given two equally good roads, the majority of drivers will pick the more crowded road...

... and of course I have fewer friends than my friends have on average.

So the following is not too surprising: We assume that we can assign a "quality of life" value Q to each world in the multiverse of all possible worlds, with Q=0 meaning "boring as hell" and Q->infinity meaning a world "approaching heaven". But then we either live most likely in a world of less quality than the average world - or the total Q grows (less than) linear with the sample size, which means the multiverse as a whole is of low quality...

inspired by Marginal Revolution