Strategic Altruism

> You identify that in fact humans have literally evolved to be cooperative and socially helpful, but you decide this is just "machievallian" and a trick of some kind?

Psychopaths trying to reverse engineer regular human behaviour can lead to funny results like this.

> The rest seems like a long walk around non-cooperative game theory in different terms.

Maybe I read it wrong, but I understand the main point of the essay is to emphasize social/cultural evolution and de-emphasize biological evolution of morality. (That there is a whole moral superstructure that doesn't come from biological evolution.)

Both can be modeled with game theory, but it is often awfully reductive. Game theory is a great source of moral paradoxes. A canonical example is "mutually-assured destruction" in a nuclear war. Another (my favorite) example comes from Yanis Varoufakis, he offered his game theory students a way to pass exam without grading, but only if they could actually act selfishly-rationally and without implicit distrust to the other members of society. The paradox is enhanced by the fact that they weren't even able to reflect on their decision. This shows the implicit power of social relationships in our decisions, but then - aren't we supposed to make free decisions?

This is also true on personal, not just societal level. With the current events, I am realizing how much of our moral dilemmas are really based on friendships and acquaintances (and you can see the difference between common folks and politicians/oligarchs). The banality of evil is that we often decide to favor existing friendships against firm moral principles.

I realized this watching Angela Collier talking about sexual harassment in academia; assuming I was a fellow professor and another professor, a friend for many years, was accused of sexual abuse, and me knowing that he has "a thing with the ladies", who would I trust, him or some student I don't really even know? Would I be willing to completely abandon the relationship in order to be completely impartial? I think even the rational Machiavellian explanation is not sufficient here (it might be actually more advantageous to be seen as impartial than to keep a friend, it's just I can just like that particular person more and therefore be more forgiving), and this shows this is not really just a status or hierarchy problem, but it permeates all social relationships.

>As a _group_ I would expect animals that abandon their young at birth to be outcompeted by those that rear them to weaning.

Actually, it is even simpler than that. Genes don't care about the unit of selection, they only care about how many surviving copies there are of themselves. People seem to have this impression that selection occurs at the level of the multi-cellular organism. Multi-level selection theory tells us that selection occurs at every level simultaneously: planetary level, species level, group level, multi-cellular organism level, cellular level, protein level, gene level. The proteins inside a single cell organism must necessarily be altruistic and cooperative. The same applies to individual cells within a multi-cellular organism. So altruism isn't a surprise, it is the norm at specific levels. Why is that the case? The answer lies in kin selection theory and the concept of inclusive fitness. The cells cooperate, because they know that the neighboring cells have an incredibly high genetic similarity. The genes don't care if they are in your liver or your brain, they just want to have more copies of themselves. Therefore if a liver cell helps a brain cell, it will automatically increase its inclusive fitness.

From the article: >However, it (kin selection) cannot explain why we are so often friendly and fair-minded towards non-relatives.

This is not true. As mentioned above, the genes do not care about the unit of selection. As long as an altruistic gene is able to detect another copy of itself in another organism, it will increase the fitness of both organisms to have that gene. "kin" does not necessarily have to relate to blood relatives at all, it just happens to be the most reliable predictor.

Considering the complexity of the human brain, it doesn't necessarily require a specific gene, it is possible that there are a multitude of behaviorally identical genes with the exact same cooperative function that happens to fool the other genes into believing that they are all identical.

The benefit of sexual selection is the ability for the brain to enter the fitness function. Instead of relying on pure chance alone, individuals can now select a partner based on their partner's genes' ability to cooperate with their own genes. Instead of a sequential chain of random mutations, it is now possible to have mutations happen in parallel with the brain acting as an evaluation mechanism. Indicating fitness has a cost that every generation must carry, but it leads to faster adaptability across generations. Poor "static" performance may be the cost for better "dynamic" performance. If a new predator emerges, female peacocks may derive information from the pattern that will help the next generation quickly adapt to that new predator.

>The rest seems like a long walk around non-cooperative game theory in different terms.

Yeah, the point is always the same: For me to know whether I should cooperate, I need to know that the other person will cooperate with me in a prisoners dilemma situation. The easiest way to do that is to have an altruistic gene interact with another copy of itself. The alternative is to have a third party hold both accountable aka "reputation management". Reputation management is an attempt at estimating group level fitness and converting it into multi-cellular organism level fitness through social interaction by doling out punishment and reward.

The problem with the classical group selection theory is that people thought it was at odds with individual selection, in a "there can only be one" manner, when in reality all levels happen simultaneously.

What seemingly is missed though, is that if there are ways to increase group level fitness by reducing and increasing the fitness of specific individuals in an unequal way, shouldn't everyone, including the punished, be rewarded for the total increase in group level fitness? Who deserves the fruit of something that was created collectively? In capitalism, all these fruits seemingly go to the top, which has a long term problem: If group fitness is not rewarded at the individual level, people will cease increasing group fitness or perhaps even decrease it to a level that is in equilibrium with the current level of individual rewards.

Pretty odd comparison as the selfish gene is about gene level incentives, not individual level ones. there's even a chapter devoted to the chromosomal effects that lead to eusociality.

then again maybe the people you're thinking about are focused on the selfish, rather than the gene.

There's part of a chapter in TSG calling out those who read the title of his book and assume that he's criticising morality/intention. All of the observable consequences of Wilson's group selection can be derived from gene selection, it's a lower-level theory (which, again, is discussed at length in TSG, as is cooperation).