OK, math.

There are 200 people in Group A. Each has the choice of taking 1 point for themselves or giving 2 points to the group. If everyone gives 2 points then everyone gets 2 points. If everyone keeps the 1 point then no one gets anything from the group and everyone has only the 1 point. But if you keep your 1 point and everyone else gives 2 points then you get 1 + 1.99 points = 2.99 points. You can obviously see how everyone might end up defecting here.

But there are also groups B and C who are all under the same rules. After each iteration everyone in the group with the lowest average dies, is ineligible to receive points in any future iteration, and is replaced in the next iteration with a new group that behaves randomly. Each individual gets to accumulate points for 100 iterations before being replaced, unless their group dies first.

Now what happens?

Without hard-coding the groups as separate evolving units, you'll have a hard time.

In general, it is easy to show the group behaviors you want if you can treat the group as atomic for the purpose of the model. But you've just moved the goalposts, your 'individuals' are not the unit of evolution any more, and you've simply set the required behavior in the fitness function.

So try it and see.

Whether this has any relevance to the real world is the debatable matter. I would argue that it does, there are many examples where strong communities have grown as a result of general goodwill and altruistic behaviours, and these communities prosper compared to those where exploitative behaviour is common. There is of course another element which a simplified model neglects - communities, once formed, exert a strong pressure on its members to conform to the norm, be it good or bad. This pressure has the effect that the individuals who depart from community norms are "evolved out" so to speak. In case of altruistic communities, this protects the group from infiltration by exploitative individuals. In case of exploitative communities, unfortunately, this perpetuates the communal problems. Eventually, if the two communities come into a conflict, the exploitative community will be eliminated.

Are you saying the simulation is not valid unless collaboration between individuals arises as an emergent behavior?

Unless you're disputing the existence of group selection in real systems, what does hard coding it change vs. a given system where it is an emergent property?

You can, of course, give a narrative argument on why you feel it is justified to exclude the interplay between these levels. But if you, at some point, try to show it is justified, by removing the assumption from your model, you'll struggle, I predict.

Evolving systems are very complex, it is an unwise assumption that you can chunk their dynamics into levels of abstraction and expect those dynamics to function independently or predictably when combined.

Now selfish behaviour kicks in as every individual wants to give as little as possible. But survivial of the group becomes crucial in face of an adversory. In that case everyone will contribute towards maximising group's strength. So in my understanding nature and wars (to a lesser degree) shaped altruism. But mostly nature. There's no way monkeys can survive without taking care of each other and group dynamics are strongest amongst them.

Altruism is completely practical.

I also agree that it is about the actions more than what you gain by doing so.

Charity is a good example of how altruism helps yourself, I donate to charity every now and then and of course it makes me feel good about donating, thus the act has an emotional gain for me.

But I don't consider it "altruistic". I help them because I want to help them and enjoy having done so, and not out of some innate sense of ethics or of what needs to be done.

I think people doing altruistic things do enjoy them, but they don't do them because they enjoy them. They do them because they need to be done, and the enjoyment is a side effect.

Of course, this distinction makes it incredibly difficult to know if someone is actually being altruistic or not.

For example, a restaurant owner might serve food for free for homeless people for increasing their own reputation and social wellbeing, vs the owner who does it entirely out of the goodness of their hearts. The net result is the same, so why worry so much about the motivation behind it? I'd say they're both equally good people, in this example.

Being able to be generous and therefore altruistic is also a decent signal to the rest of your group you're fit and healthy.

What is the difference between the two conditions? The only one I can think of is that removing the side effect does not change the behavior, but removing the main motivator does. How is this significant though? Is it easier to lose enjoyment from altruistic activities than to lose a sense of responsibility? Is pleasure more fickle than duty?

That distinction, IMHO, also makes the definition of altruism so narrow as to be useless.

I think that I might go as extreme as to say it is only altruistic if it makes you feel bad or puts you in a clearly bad position.

Let's say that out of an act of altruism you help a person achieve some goal of theirs. Now you do not have a material gain from this, but doing so will make you feel good, hence you did it to make yourself feel good?

What you described there is in nature called symbiosis [1]. Like a cat or dogs lives with a human, or an ant with a fig tree.

[1] https://en.wikipedia.org/wiki/Symbiosis

Artificial studies like this have practically no bearing on actual behavior of real populations in real situations. Try the same study with a family, or a football team or the employees of a company or a military unit and then see what result you get. That's what reality is like. Real people don't live in randomly selected study groups, they live in communities.

Maybe an additional rule where the 20% with less points of the population of the groups with more points also dies. So there is actually an incentive for the individual that keep the points.

* Divide the big group into smaller groups (where they are giving each other inside the group only).

Now individuals in the group where many give to each other get an advantage over individuals in other groups.

Individuals that keep for themselves while staying in a group that gives also keep the better advantage on cost of their group.

* Allow the group to punish those who keep for themselves.

We are not ants.

Now in emergency situations, my experience is that we ARE hardwire to help each others. It works almost automatically, and is quite magical.

But as son as the immediate threats wears off, all the stuff I mentioned come into play.

https://en.m.wikipedia.org/wiki/Tit_for_tat

http://www.robertsapolskyrocks.com/behavioral-evolution.html

I think that this was easier explained in "The Selfish Gene". Individual altruism is an emergent characteristic of a selfish gene.

I risk my live 10 times to save others until I die without off-spring. It is altruistic for me, but the gene has saved 10 people that are probably related to me. That means that the gene has a probability over 10% to be in that other individuals it is a win for the gene.

From my point of view it sucks, I don't have offspring. But the gene survives and prospers. This is still more extreme in species like bumblebees where sisters share 75% of their genes.

For monozygotic twins there is a similar case. From the gene perspective is irrelevant if one survives or the other, even that it is really important for the individuals.

- It's easier to find a significant other with whom to produce offspring if you exhibit kindness and charity to others.

- Generosity is often reciprocated.

- Giving implies EQ so altruistic person also likely takes care of family.

That's the issue with statistical models with finite decisions and states, they don't include enough parameters and interactions to capture anything relating to real life.

Like imagine a prehistoric human feeds a weak stranger, later that month the stranger warns them of danger. It's so obvious how kindness could be beneficial gene.

E.g. suppose I have a loaf of lembas. I can eat it myself, and it increases my chance of surviving and reproducing merely by 10%, because, while I'm starving, I'm healthy. But I have three siblings who share 50% of their genes with me, and who are starving and sick. If I give all my lembas to them, their chances of surviving and reproducing increase twofold, and become same as my current chance. In that case, it makes sense for me (or rather my genes) to evolve the behavior whereby I give away the lembas, because these three survivors will spread the genes that we share more efficiently than I alone.

Now, obviously, the math here has to add up in real-world scenarios - is that what you mean by "it wasn't trivial to stack the dice"? It just feels like it shouldn't be hard to come up with many cases where all the coefficients line up...

"No, but I would to save two brothers or eight cousins."

As you share on average half your alleles with a brother and one-eighth with a cousin, Haldane was giving the number of relatives one would have to save to "break even".

As quoted in Mathematical Models of Social Evolution : A Guide for the Perplexed (2007) by Richard McElreath and Robert Boyd, p. 82;

https://en.wikiquote.org/wiki/J._B._S._Haldane

This is in relation to the idea of kin selection: https://en.wikipedia.org/wiki/Kin_selection

There are instincts that become ingrained as part of evolutionary process, and they are those instincts that correlate with genes that propagate more.

Thus, the genes that you inherited from your parents may activate an instinct that will make you sacrifice yourself (or at least reduce your chances to reproduce in the future), so long as that results in a larger portion of those inherited genes in the gene pool down the line - e.g. because some of your siblings, or even more distant relatives, survived to carry them on.

It's one other reason why "survival of the fittest" is really a very misleading way to put it - it's about "survival" of a particular package of genetic traits, not about survival of the species, much less individual specimen.

Evolution is a game of information propagation, in which actual living beings are heavily parallelized delivery channels, information transmitted is the configuration of channels, and there's non-random noise (environment) affecting signal propagation through those channels. The "winning" configuration is likely to involve some channels that do not transmit data directly, but are dedicated to performing error correction etc for other channels, that do transmit information about the entire configuration.

In fact, over time, you may even end up with a configuration that can dynamically allocate channels for error correction, if and when the noise becomes noticeable enough (i.e. environment gets more hostile), so as to not waste valuable bandwidth when it's not necessary. And then we call such things "altruism", "heroic sacrifice" etc.

The principle behind this is that evolution operates at the level of genes, rather than individuals. I recommend Richard Dawkins' The Selfish Gene for a fuller explanation.

Additionally, when at large scale war, often you can trust your cultural bretheran better than the enemy. As a result, its important to have as many healthy and proud people as possible inorder to go to war. A culture of kindness and giving can support that.

Additionally, this is all about genetics. While there is always an argument that your genes are the best genes, cultural agreement can give faith that the best will rise and flourish. Treating your neighbor as you would yourself reinforces the hope that even in failure of the individual, the group can still succeed.

Im sure theres more

Imagine a game you play with someone where you have 40 units and you need to divide it between the two of you. The other person has the choice of accepting the deal or refusing the deal. If they refuse, both of you get nothing. How do you divide the 40 units so that the other person accepts the deal?

Naive economics points that even if you offer the other only 1 unit, the other should take it because if they refuse the deal, they wouldn't have even that unit. In practice however, people down't work like that. If people feel they are being treated wrong, they tend to go for lose-lose. And so, in these kind of games, most people go for a fair split 20-20.

Now, imagine the same game but with a twist. You have a bunch of instances of these games and your objective is to have the highest score across all the instances. The trick is that the other player has to chose you to play with. What strategy would you adopt so that you maximize the score?

In this game, a 25-15 split is better. You lose the small game BUT more players want to play with you and so, you win the BIG game. Altruistic players are better at playing the BIG game, the long term game.

If you ask this question, you either have to have one example where there is no payout at all (note this should be an example most people agree on) or you accept there is no altruism, at least not following your definition of the word.

It seems to me that their interpreters often assume that identifying some behavior as evolutionary advantageous to an individual or group is somehow correlated with the goodness or desirability of that behavior. That's a fallacy. Whether the behavior is commendable, good, desirable, etc. is not related to its capability of increasing the long-term survival and spreading of someone's genes. (I'm not claiming everyone makes this association, but some seem to make it.)

Moreover, there seem to be countless examples of behavior that we are perfectly capable of and whose effect is difficult if not impossible to explain with evolutionary arguments. Some of them are considered commendable, other not. Random examples:

- spending your life in almost solitary confinement to solve some math problem

- deciding not to have children

- waging wars

- not passing your wealth on to your useless son and instead give everything to a charity

- being gay or asexual

- writing a novel without getting paid for it and that also doesn't increase your mating chances

The list could go on and on. Of course, evolutionary explanations may be constructed for any of them, but it's kind of doubtful that all human activities fit into this scheme. So my two points are:

a.) Whether altruistic behavior is good/desirable or not has nothing to do with its contribution to evolutionary fitness.

b.) Within some limits, we are perfectly capable of behaving in ways that are not evolutionary advantageous to us.

Finally, evolutionary models of social behavior that are not based on the genetic transmission of behavior simply do not have anything to do with real evolution. I've often seen real evolution being deliberately mixed up with ideas about the spreading on information contained in books, web pages, speech, etc. but I fail to see how the spreading of this information could be even remotely related to the way genes spread. Once you hear about a social fitness function or the fitness of a meme, you can in my opinion safely assume that the author pulled that concept out of his ass or copied it from others who did so.

Behaviours that groups exhibit will definitely be passed on from individual to individual and definitely have an effect on the ability of the group to survive.

And - whether something is passed on socially or through genes, it doesn't matter. 'Evolution' is not just a function of genes, obviously.

Your list of 'behaviours' is easy to describe, and I'll do it in context of your 'a' and 'b' positions.

From 'b' - "we are perfectly capable of behaving in ways that are not evolutionary advantageous to us" - obviously. But who cares? The species that takes up the best behaviours for it's environment will persist beyond those that do not.

From 'a' - Whether altruistic behavior is good/desirable or not has nothing to do with its contribution to evolutionary fitness. Yes it does. Run a little evolutionary simulation a a computer program. Create little 'bits' that evolve over time, and interact with one another - you'll see the bugs with the best social behaviours win - the others dissapear.

- " spending your life in almost solitary confinement to solve some math problem"

First off - this did not happen until way, way into human social development. Once we have complex social functions, it makes evolutionary sense to allocate some members of society towards 'very long term' thinking. 'The guy in the cave' is the same thing as a PhD today: someone who'd dedicated their life to something probably trivial or irrelevant, but which may have huge impact, possibly in the future.

- 'war' - war is perfectly understandable as selfish behaviour in the short run: a social group can 'get ahead' quickly by taking stuff of others.

- 'not passing on stuff to kids'. From a social perspective, it doesn't matter that much if the money goes to one's kids, or the neighbors kids. Both will contribute to group survival.

"I've often seen real evolution being deliberately mixed up with ideas about the spreading on information contained in books, web pages, speech, etc. but I fail to see how the spreading of this information could be even remotely related to the way genes spread"

It's very simple: an organism that learns how to read and write, and to pass down it's knowledge from generation to generation will 'out survive' ones that don't - and their genes are much more likely to be passed on.

That's not obvious at all and it matters a lot. The idea that there is e.g. an 'evolution of ideas' is ill-conceived and misleading for many reasons. The same can be said about similar ideas like 'evolution of social systems', 'evolution of memes', etc.

In order to show that there is an evolutionary process that goes beyond of what is transmitted by genes you first have to show empirically that the respective information is transmitted non-genetically, then find a good model of how exactly it is transmitted and to which individuals and groups it is transmitted, and then develop and show that a concept of fitness applies to the process that is sufficiently similar to biological fitness to even be able to talk about evolution in a meaningful, non-contrived way. Then you can attempt to build an evolutionary model, and then you'd need to evaluate that model by looking at predictions. Unless the information is spread exactly like genes are spread, that model will look very different from biological evolution and each and every aspect of the model needs to be justified, in the same vein as you have to justify any analytical model.

I don't claim to be an expert in this field but have to reiterate that from what I've inferred from many talks and conversations with colleagues who work in such areas (like EGT, social dynamics, etc.), that there is no widespread consensus on such a justification. Most models of non-genetic, nonbiological 'evolution' seem to be fairly ad hoc. They just stipulate certain information transmission mechanisms and come up with fitness criteria as it suits them, without much of an empirical basis. It's Lamarck's giraffes in fine modern mathematical clothing.

From 'b' - "we are perfectly capable of behaving in ways that are not evolutionary advantageous to us" - obviously. But who cares? The species that takes up the best behaviors for it's environment will persist beyond those that do not.

In my experience, everybody cares about what's good and what's bad, even those who think they don't. Not caring about it and placing evolution higher in your value system is a bit like saying "Hey, let's blow up the earth, that's more consistent with the laws of thermodynamics anyway."

Run a little evolutionary simulation a a computer program.

With arbitrary parameters, fitness function, and information transmission mechanism, I can certainly write seemingly convincing little computer programs to show of any behavior whatsoever that it is evolutionary advantageous or disadvantageous. That's because if I can choose all parameters and the transmission mechanism as I like, I simply redefine what 'evolution' means each time, disregarding how biological evolution works. But just because you make a dynamic model with some sort of fitness function, many individuals, and some starting state doesn't mean you model evolution.

It's very simple: an organism that learns how to read and write, and to pass down it's knowledge from generation to generation will 'out survive' ones that don't - and their genes are much more likely to be passed on.

I'd like to see evidence for that. But in any case that would only explain how reading and writing evolves, not any evolution about what's written - which is what I was criticizing. The written information is not transmitted by genes, hence not subject to biological evolution. If the written information is somehow subject to some evolutionary selection process, then you may try to find an analytical model for that. I'm not saying that all of these models are wrong for some principle reason, just that many of them are fishy, based on sometimes flawed analogy to biological evolution, and in need of more foundational justification.

I also kind of disagree with many aspects of your 'easy' explanations and would like to see some evidence for them, but let's leave it at that.

Where did you learn to speak English?

Where did you learn to read and write?

Where did you learn Math?

Where did you learn how to eat, run, catch a ball?

You learned those things from your social environment - those are bit of knowledge that are handed down from generation to generation and 100% affect your ability to survive and thrive in your environment.

You're a living example of how 'information is passed non-genetically' - there's no need to do an experiment to show how that is done, it'd be like doing an experiment to see if the sun rises tommorow.

Social information is absolutely passed down from generation to generation, and this affects how species adapt to their environment and thus which genes are passed. This is obvious, there's no way to contend that, really.

Even if we acknowledge that some of those examples can't be explained by evolution, evolution permits aberrations. It's in the very core of the model.

The others are indeed examples of memes overpowering genes, and an evolutionary model (of memes) can help to explain them as well. It's not a well-developed field to be sure, but it's a worthwhile perspective. A cartoony example is religions that ban sex vs religions that ban contraception - given that religious memes are strongly conserved in offspring, which meme would you expect to spread better? How does that compare to what we actually see?

Given homosexuality is prevalent in the animal kingdom as well, that's pretty much assured.

Altruism is one of the most pervasive forms of interaction in life on Earth because it is the mechanism by which multi-cellular life functions at a basic level.

It's really more about deception than anything else.

That said, I do expect it to be hard. Nature created cooperating species more than once, but not that many times¹. And it had half a billion years of a simulation that was certainly much larger than yours.

1 - Counting placentarial mammals only once, what is very likely correct.

When I see how much investment we take in people these days, it comes naturally to me that we have to look after each other.

People are expensive to produce and maintain and experience doesn't grow on trees.

Assyrian empire was extremely brutal to people it conquered, Persians who took its place was more "humane" to their conquered subjects as a policy. Resulting in a relatively more stable land and subjects were submitting "willingly". From that point on, it was their main selling point.

https://en.wikipedia.org/wiki/Price_equation

https://en.wikipedia.org/wiki/The_Evolution_of_Cooperation#A...