Thanks, I've made this edit—the capitalisation comes from the original where it's more cartoonish and seems less shouty, but here, yes unintentional shouting :)

Thanks Arturo, and yes, I found it very interesting. I find the passage you've pulled particularly interesting in terms of AGI which seems to be following the same notion of "de-instinctivation" substitution narrow hardwiring for the ability to adapt to general situations.

I also found the critique of Utilitarian impartiality interesting, and it helped clarify my thinking in terms of the simulation. This model actually is founded on cognitive bias, treating it as a feature rather than a bug. So, incoming beliefs are only ever adopted when aligned with pre-existing beliefs and yet it is still impartial in that, given a conflicting belief an agent will always choose the strongest coalition of beliefs, letting beliefs with negative marginal value go freely. This is actually a very reasonable way to adopt beliefs (because we can only ever make decisions based on our prior knowledge) and yet has the natural result of bias that we observe in humans.

Thanks for providing that link Nathan, that does seem to significantly undermine the Mani et al study. While I agree with you that at this point primary studies should be, by default, seen as not credible, it did help a lot (in terms of convincing me personally) to see a study specifically designed to replicate the Mani et al study.

Do you find any evidence for the conclusions in the post credible? Or are you aware of more credible studies that would support the argument the post makes? You seem to know your stuff, so, I don't want to waste your time, but I would value your input about whether the post's position is tenable at all given the available evidence.

I don't want to be posting nonsense, so depending on the evidence available I would either rewrite it with more reliable evidence, or take it down.

Thanks for your comment. I've been aware than this perspective is prevalent, but I haven't actually seen examples of where replication of the same study has been attempted, I have only seen some that introduce other major factors that one would expect to influence results. The link you sent me to criticises priming in a broad way, pointing to heuristics like the effect being too large to be believable, which seems a pretty subjective judgment.

The link specifically criticises Danny Kahneman for using priming in small studies to make large generalisations, and in Kahneman's response he makes a fairly good rebuttal. The one thing he concedes is the small size of the studies he used, which is not the case in the priming case used in this post, which involved a series of studies with several hundred participants each.

I appreciate that I might be incorrect to have confidence in these studies, in light of the widely held opinion that priming studies are not reliable, but I'm yet to see the specific studies that have attempted, and failed, to replicate these specific studies.

Hi Nathan,

Thanks for your response, and I see your point, the more specific the parameters get, the less general the conclusions can be.

To explain, my purpose in using a simulation is to illustrate a phenomenon that is perhaps too complex to reduce to a formula, because it seeks to emulate some aspects of society that are often not accounted for in game-theoretical models. Simulations allow for complex parameters to provide a sort of empirical evidence for principles that might not be able to be proven mathematically (by me at least).

The reason I've chosen the parameters I have, is not to create an inevitable outcome, but to reflect aspects of the real world that are not usually considered in game-theoretical models, like for instance the instinctive animal behaviour to avoid agents with whom you've had previous negative experiences. This is difficult to model mathematically, but is never-the-less a significant factor when creating a model that applies to the real world.

The stipulation that no one dies is a simplification that serves two purposes:

1. It reflects the self-balancing nature of human systems in the simplest way possible (modelling death and reproduction, social welfare, health and other systems that maintain a society's population would be unwieldy—and even more specific).
2. It gets the population past the inevitable point of total collapse that happens if the defectors are able to essentially kill other agents during their dominant phase.

So, the specifics of the model are not meant to be arbitrary, but reflective of features of actual populations of people or other animals. The aim was to better approximate real world dynamics rather than the siloed game-environments which often result in conclusions that don't comport with common sense—not because common sense is wrong or the theory is wrong but because the game-environment is too limited.

Your point is important though, and if I develop this further I would think about introducing controls for the initial ratio of the agents (cooperators:defectors) and less specifics preserving survival, so collapse becomes a feature determined by the ratio. Other controllable parameters might also help to give the user a more intuitive feel for the effects of various dynamics on the system.

Thanks, I think I may have used the wrong term "source"-based, meaning "origin" or "residence"-based. I was meaning that generally a company is taxed in the country where their offices are, not where their consumers are.

That was refreshing change of pace :) Thanks for posting.

Great article! This seems like a valuable avenue for people with the relevant skills.

I’m interested in your observation that poorer countries tend to prefer a source country-based tax, while wealthier countries lean toward a consumer country-based tax. However, it seems paradoxical that OECD countries, which hold stronger influence over international tax law, have led us to a predominantly source country-based tax system. Could you comment on why think this is the case?

I’ve been interested in the idea of a consumer country-based tax ever since I worked on a film with Thomas Piketty, which loosely covered his Capital in the 21st Century. If I understood his view correctly, a consumer country-based tax could potentially help eliminate tax havens because companies can’t control where their consumers are in the same way they can control where their offices are based.

I'd love to hear your insights about this.

By donating the value of my scholarships, I have likely impacted 50,000 lives, prevented 9 deaths, and significantly improved the lives of thousands of animals.

Numbers leave some people cold, but as with most EAs I'm a big believer in the power of numbers like these, they make me feel warm inside. Thanks for sharing your cash, but also your story, as a fellow pledger, I hope these stories inspire others to do the same.

What are your thoughts about the position of existential risk—and more specifically AI x-risk, as a key concern of Effective Altruism?

I have a sense that while it is a significant and important concern, I'm not sure it falls into the category of "altruism" as opposed to "self-preservation", and considering its current popularity, is there a risk of this concern crowding-out other core altruistic causes around the immediate well-being of those less fortunate or less empowered?