Values Darwinism

post by pchvykov · 2024-01-22T10:44:41.548Z · LW · GW · 13 comments

Contents

  TL;DR / summary:
None
13 comments

TL;DR / summary:

Can we objectively define "progress"? A Darwinian definition seems plausible. But when referring to cultures, world-views or theories, it seems that physical survival of the fittest must be complemented by memetic survival. The interaction of these two will generally be complex, and may parallel models from epidemiology. Most discussions usually account for physical or memetic selection - but not their interplay. If se suppose that happiness makes memes more contagious, we can push back on the common idea that "evolution does not select for happiness," and update our conventional notion of progress. 

[Cross-posted from my website http://pchvykov.com/blog]

---

I’ve had some discussions recently about how to measure whether our conventional idea of “human progress” really constitutes an improvement. There is some evidence that people were happier and more well-nourished in caves than directly after the agricultural revolution (see Sapiens by Y. Harari). Now there are ideas that “the simple life,” like at a permaculture village, is somehow more conducive to our humanity and well-being than city or corporate life. We could make the argument that the question of "progress" reflects the more fundamental question of “finding deeper truths”: On the one hand, we tend to think that our modern scientific understanding is “more true” than e.g., Aristotelian physics, or Buddhist cosmologies. On the other hand, we understand that "all models are wrong, but some are useful" (G. Box) — and perhaps some Buddhist ideas are more useful if the goal is human happiness (e.g., see Bhutan's Gross National Happiness). Note that for this and following discussion, we entertain the scenario where we do not have access to any fundamental notion of “truth” that we could universally agree on.

One argument that always seemed most convincing to me in favor of “progress is real” is Darwinian one: ultimately, success should be measured via “survival of the fittest.” Even if hunter-gatherer tribes were somehow happier, they were outcompeted by agricultural communities, which could support much larger, albeit more miserable, populations. Similar argument can be made for modern science: technological progress gave Europe the firepower to subjugate most Eastern nations. This idea could be summarized by refining Box’s quote to be “all models are wrong, but some are useful for long-term survival of their proponents” — and those models will be the ones we will see, get used to, and think of as "best" or "truest."

However, while Europe may have “conquered” India in physical space, lately it has become hard to ignore the proliferation of Yoga studios on every corner of major Western cities. This raises the idea that besides “physical Darwinism” – survival of the fittest physical organism – we must also consider “memetic Darwinism” – survival of the fittest idea. “Memes” (see Memetics) are ideas or “cultural units” that multiply, proliferate, and mutate against the substrate of the network of communicating human minds — not unlike epidemiology of diseases. Thus, while you are physically fighting and defeating another culture, that culture may be "infecting" yours with its ideas or world-views. Ultimately, the long-term survival of an ideology will depend on both, the rate at which its proponents physically survive in a competitive environment, as well as on the rate at which it can “infect” others’ minds (or be “displaced” in the minds of its proponents by other ideas). Note that this is a qualitatively new dynamic compared to Darwinism in the animal kingdom — where most non-human animals’ capacity for adapting fundamentally new behaviors in a single lifetime tends to be more limited.

As such, we can once again adapt Box’s quote to be “all models are wrong, but some are useful for their own long-term survival” – i.e., we get “model Darwinism.” What then determines “model fitness”? This question is especially interesting if we want to define our notion of "progress" (or even "truth"?) in terms of this fitness, as suggested earlier. Indeed, how does this fitness relate to our conventional or scientific ideas of truth? I think these relationships may be hard to guess, but could perhaps be quantitatively researched. We might hope that models conventionally seen as “closer to truth” give either a physical advantage (e.g., modern science) or are more contagious (e.g., yoga). Nonetheless, given the Darwinian success of astrology and tarot reading, it seems that the connection is non-obvious. In particular, it's clear that model fitness alone cannot be used to legitimize the notion of “technological progress.”

It could be interesting to leverage the research of epidemiology on networks to study similar trade-offs between physical and psychological effects of different world-views. E.g., we know that diseases that kill quickly often do not gather a high death-toll as they have little chance to spread. In model Darwinism, this could be paralleled to suicide cults. Modern science could then be compared to something that makes you live longer, but is not very contagious (though arguably this is starting to change). In contrast, more contagious ideas may perhaps be correlated with ones that give a “psychological advantage” – such as for subjective sense of well-being, purpose, or belonging. If an idea is physically helpful, but extremely distasteful, it will become unstable as it is easily displaced by more contagious ideas. This could motivate an interesting new way of modeling burnout, or conversely, altruism (see also my post on the Paradox of Tolerance).

Conventionally, we often think of material advantage as the sole criterion for what makes an world-view "good" or "progressive." But we are not purely physical beings. Long-term misery may be selected for by physical Darwinism, but not by memetic Darwinism. This way, we may start to push back on the idea that "evolution does not select for happiness" - which may be an artifact of ignoring the memetic contagion effects. If the trade-off is between long and miserable life, or short and happy one, the choice may be individual, but a bilateral view of evolution as presented here would likely select some balance.

13 comments

Comments sorted by top scores.