Biological evolution selects roughly for creatures that do whatever it takes to have more descendants in the long run. When such creatures have brains, those brains are selected for having supporting habits. And to the extent that such brains can be described as having beliefs and values that combine into actions via expected utility theory, then these beliefs and values should be ones which are roughly behaviorally-equivalent to the package of having accurate beliefs, and having values to produce many descendants (relative to rivals). Equivalent at least within the actual environments in which those creatures were selected.
Disturbing isn't dystopian; I agree this isn't so bad.
I agree that, eventually, evolutionary dynamics seem likely to produce beings that consciously try to optimize their long-run evolutionary fitness.
But I would caution against describing this as simply "more descendants". With traditional genetic evolution, for example, the key metric is inclusive fitness (which includes related individuals that are not descended from you; see also kin selection). What metric is relevant for future, non-genetic evolution depends on how reproduction works.
Also, I don't receive this as negatively as you seem to. All our current instincts ultimately came from natural selection, after all, which includes the ones we like, not just the ones we don't. So far, social systems claiming to be based on evolutionary ideas have a deserved bad reputation (e.g. "social darwinism"), but they had quite a shallow understanding of evolution and evidently were not actually successful at long-run evolutionary fitness.
I'd guess that an individual optimizing for long-run evolutionary fitness would act something like Elon Musk: both accumulating material resources and reproducing a lot. And, though Musk is a flawed human being like the rest of us, I don't see a future of people-similar-to-Elon-Musk as a dystopia.
Group A is doing much better than Group B currently. If group A and group B were in direct conflict with each other then I would expect group B to dominate in most cases. But if they were to be allies, I would expect group A to dominate.
Holding expected life expectancy constant, descendant maximization necessarily needs to go up from its current values in all of the developed world (Group B) because we are below replacement levels.
Selection becomes more important if we reach max resource capacity and each individual birth depletes more resources than it generates (not effective enough).
If individuals are created by entities other than themselves, reproduction would not necessarily be an important value for such individuals. The polis entities in Greg Egan's "Diaspora" would create human-similar AIs except the AIs would be designed to be disinterested in reproduction, to avoid creating internal conflict within the polis. We could evolve to be eusocial, with suppressed individual reproduction and with groups becoming persistent over cosmic timescales and capable of non-evolutionary self-modification.
The evolution of humans so far demonstrates the opposite trend regarding which values emerge. Evolution favors maximum descendants indeed. It is indifferent to the interests of an individual, and often directly contradicts them, for example, anger and extreme aggression. The necessary traits get encoded as instincts and emotions, and that's all that animals know, they cannot help but follow that.
But humans developed the ability to suppress instincts and emotions in favor of personal interests. This is often detrimental to maximizing offspring. And I put it to you that humans are winning evolutionarily compared to animals.
You are suggesting that evolution will make us like sacrificing personal interests for the common good or for the good of the descendants. But this seems backwards, the opposite of what happened so far. We used to fight and die to gain, say, women when it was very risky. Now we are capable of deciding not to do that because it is better for our individual survival, even though it's worse for leaving offspring.
The persistence of territorial animals demonstrates an alternative possibility. We should expect most future agents to be small agents that care mostly or entirely about replicating themselves, but most future territory (i.e. productive resources) to be controlled by agents that care mostly or entirely about controlling a lot of territory.
Proletarians make more proletarians and become more numerous, states compete over territory and become larger. It doesn't seem very horrible for future Jupiter brain utility monsters to have lots of gut flora. Income and wealth are already approximately Pareto distributed.
It has always been the case - and in many ways.
Atheists tend to hedonism which used to mean not looking after children so well, and more recently not having children. Religions often have rules explicitly designed to increase child production and care.
Religions have understood the practical century to century aspects of evolution for a long time, just not the scope of it from a natural history point of view.
Fascists also explicitly took evolution as a justification of trying to be the "fittest". They should have taken more notice of religion and realised that being overly agressive about things does not increase your fitness due to making too many enemies. You have to exist with the world.
I don't think it's a common for a species as a whole die off due to overpopulation. Particularly if resorting to cannibalism is an option.
I don't think I share many values with my simian ancestors either.
Imagine the following scenario:
A population is composed roughly evenly of two groups of people (A&B)
Group A is extremely concerned with maximizing descendants (10/10) but is very ineffective at accomplishing their goals (2/10).
Group B is only somewhat concerned with maximizing descendants (4/10) but is very effective at accomplishing their goals (8/10).
At the beginning, as A & B are roughly equal the AVG amount that maximizing descendants is valued is 7/10. Over time, Group B consistently outcompetes Group A simply by being effective, resulting in a population that is dominated by Group B. This new population values descendents at about 4.2/10 (it is 95% composed of Group B now).
The point is to demonstrate the idea that even with evolution in play, descendant maximization can still decline as a value if there's even 1 separate thing that pushes selection more (and i certainly think "being effective" is the sort of thing).
I suspect that if we could somehow achieve a calculus of fitness, "valuing descendant maximazation" would probably not be the #1 most relevant input, and certainly not the majority.
I'm not saying "descendant maximization" won't go up over time it just seems obvious to me that such a claim is far from certain
I think you're just saying that evolution is a long game, and single time period gains is not the metric. You want gains you can re-invest / compound over infinitely long periods of time. Reinvestment risk is particularly tricky over generations, particular if your adversaries can change the rules or the environment.
Building a bridge might be good for this epoch, but if you can't reinvest your advantages from it into more bridges or other things for your coalition then you might losing in the long term.
I think that's generally true. Stable evolution to care most about building stable systems of cooperation that can't be easily hijacked (aka prevent viruses & cancers). People are willing to safeguard the community, rules, relationships, the support structure, fairness, to make sure that everyone can reinvest on reasonable terms on an ongoing basis, even to the detriment of individual gains in this epoch. It's a more advanced "Live and Let Live" equilibrium that secures peace.
It's the over-simplified notion of "disruptive" or "winner take all" competition that disregards this, and only looks at single time period gains, and concludes everyone should be willing to sacrifice everything around them for those simply accounted-for present gains. That leads to total war as everyone tries to seize the strategic high ground of disrupting future reinvestment by their rivals. Everyone competes to hijack the system, rather than competing within it.
Your partial solution of stopping evolution really only needs to apply to the evolution (an arms-race) in hijacking / viral abilities. Stopping evolution of any sort is difficult, but a big part of the evolution of "civilization" is everyone recognizing the need to pitch in to do so as part of their "individual", long-term, interest.
In the last 20 years, people have become willfully blind to this kind of reasoning, as some sort of starting gun was fired for a Great Looting Race, to squeeze all the present gains they can from the current system before it's gone.
Self-directed evolution. Design your descendants. You can't stop natural selection, but you can change the selection criteria and create your own macromutations. So: want a future in which people are deluded about their motives? We could make it so. That seems to be what has happened so far, anyway. Our motives are base and ugly. However: we can paint a picture for others in which we are holy saints - and ideal business and life partners. If we even believe it ourselves we will likely be all the more convincing.
If you look up the definition of the term "evolution" in textbooks on the topic, you see it means something like: "changes in the frequency of inherited traits in a population over time". It doesn't typically specify the origin of the changes. Random mutations, reinforcement learning, genetic engineering, intelligent design or whatever - any kind of inherited changes are forms of evolution - according to standard definitions of the term.
Rot is hard to avoid, even with our tech: https://www.overcomingbias....
Yes, but what is hard for biology and what is hard for the human technology stack are not that correlated. Our technology is amazing at transmitting and recording information. Those seems like the technologies that would be necessary for extremely long lived and large beings.
Biology also had that option available to it. The reason it didn't choose it is because it is very hard to successfully implement.