Effective Altruism

Arg! This post by Matthew Yglesias exhibits perfectly the problem I have with EA. I was first triggered by this:

The classic argument here is that if you were walking down the road and saw a child drowning in a pond, you’d jump in and try to save him. And you’d do that even if you happened to be wearing a nice shirt that would be ruined by the water because saving a child is more important than a nice shirt. So how can you in good conscience be walking around buying expensive shirts when you could be giving the money to save lives in poor countries?

So, this rhetoric completely ignores the “moral scope”, in scare quotes because I’m using my own private language, here. Walking near drowning children and the shirt that’s literally touching your skin is very local, whereas people living & dying in other countries is very non-local. It’s a false equivalence. But then Matthew rescues his rhetoric later:

We are not 100 percent bought in on the full-tilt EA idea that you should ignore questions of community ties and interpersonal obligation, so we also give locally in D.C.

EA’s tendency to think long-term and broadly scoped is fantastic. However, a tendency I’ve noticed in my professional work as a simulationist^1,2 is that the effort to generalize/universalize tends to also flatten the discourse. At work, the primary offenders are mathematical modelers. I don’t need to lay it all out, here, because you can simply think of Maslow’s: to a person with a hammer everything looks like a nail. Systems of differential equations are like your well-worn hammer that fits your hand so well you forget it’s even there. You are the hammer. The hammer is you. You forget that the hammer is not you. I.e. you flatten the scope.

EA does this with AI risk and people in the far-flung future as well as with currency-based donations to people in far-flung places. That’s what currencies allow us to do. That’s what “currency” means. But currency is not a crisp replacement for inter-scope interactions. We see this all the time with huge donations and, e.g., pallets full of cash shipped to places like Iraq after we destroyed their communities.

Pfffft. Forgive me if I’m skeptical of all the high falutin’ yammering about “consequentialism” when the actual, richly detailed hairball network ecology of consequences are ham-handedly squashed down to money by the “earn to give” mantra. Anyway, Yglesias rescued his rhetoric there. But then when he fawns all over a billionaire who extractively sucks money from unwitting MLM victims in the cryptocurrency game and then applies that money to pandemic prevention, Yglesias returns to wallow in the flattened discourse again. Yeah, that’s the ticket. Suck money out of all those idiots trading shitcoins so you can (arrogantly) redistribute it to your pet causes. That’s ethical, right? That’s consequentialist, right?

Now, OK. I’m being harsh. I actually do believe there are valid use cases for distributed ledgers ([sigh] “blockchains”). I still have hope for distributed storage and distributed computing. But currency ain’t it. I suspect (proof of stake) cryptocurrency will be part of our future in some lasting sense. But it won’t be what the crypto crowd says it will be … especially not what the exploitative crypto-Billionaires say it will be.

But that’s the problem with EA. In the spirit of greenwashing, EA is altruismwashing. They’ve bought into their own bullshit. And while bullshitting other people is bad. Bullshitting yourself is Evil.

1. Ören, T. The Many Facets of Simulation through a Collection of about 100 Definitions, 2011. [SCS]
2. Simulationist Code of Ethics from the NTSA.

Advertisement

On Prediction

I’m giving Substack a try.

This article demonstrates the flaw in libertarian thinking quite nicely, I think. It’s especially useful because Hanson is a solid thinker and a good writer. Please read the article rather than rely on my opinion of it. But this quote goes to the core:

As I’ve discussed before, it makes much more sense to value life-years lost, relative to income.

The association of income with value is questionable, at best. Now, it’s possible that I simply haven’t read enough and income does turn out to be a credible indicator for value and, where it’s not, there might be methodical adjustments and fudge factors that compensate. But such methods involve the layering of models. And all models are wrong. When wrongness is layered atop wrongness, it becomes ever more difficult to remain unbiased.

We can see Hanson’s bias more clearly in this post. Again, please read it. The focusing quote is:

In particular, “free” schools, libraries, and hospitals only pass this cash gift test if there is some positive externality with their use to counter the fact that making something “free” induces wasteful over-consumption of more than they would consume they had to pay for its real marginal cost.

The problem, of course, is that it is ALL externalities. Stanislaw Ulam is famously attributed with saying “Using a term like nonlinear science is like referring to the bulk of zoology as the study of non-elephant animals.” I think we can reasonably swap “nonlinear science” with “externalities” and the sentence will still ring true. More seriously, though, there are 2 problems with Hanson’s assertion: 1) on what basis do we determine a commons’ proper consumption? I.e. how do we define over-consumption of things like libraries, hospitals, national parks, etc.? And 2) if we reduce the high dimensional space of externalities down to the same measure we use for the core variables in the model (“cash”), how do we estimate the fidelity of that reduction? For example, libraries are used for more than free access to knowledge. They’re also used to warm up on a cold day living on the street. Or checking the job boards on the internet. Or dropping off your election ballot. Emergency rooms are used as an inadequate compensation for a lack of health care services. Etc. This idealistic, rationalist, cash-test fails even the first pass credibility test. But, again, Hanson is a good thinker and writer. Much of what he says about how we can translate value-thinking into policy-thinking is useful.

cults, again

So, on the heals of the premier of “The Vow” and the recent uptick in QAnon news and having recently been awed by these episodes, I feel like I’ve missed some arching narrative … like there’s a story that would make all this hang together nicely. My typical whipping post is population density and bumping up against the earth’s carrying capacity (at least in terms of our current technology). That rhetoric would play out similar to my thoughts on criticisms of “late stage capitalism” and neoliberalism’s need for an underclass. Where caste systems are normally thought of in terms of a materially closed underclass, neoliberalism allows for a dynamic, materially open, underclass, preserving some closure in one or more of the other 3 types of cause. In this caste rewrite, “the system” is under stress but is dynamic and adaptive enough to swap one underclass for another. Back in the context of cultish thinking, “the system” is under stress and everyone is looking for some (any old, most available) grand narrative to explain whatever it is they think they see.

What I’d like to think refutes my population density explanation for cultish thinking. What I’d like to think is that this has always happened, that any animal with a large executive module, regardless of the contextual details, will tend to home in on the most available narrative … inference to the best explanation, as it were. And that the Synanon collapse was in the news right as I was entering into my fascination with the occult and mind-bending near-true ideas without me knowing about it, argues in favor of the idea that it’s always been this way. Maybe the hostage crisis or Unabomber activity drowned it out. But I would have been pretty interested in a nearby rehab cult at the time.

But sheesh, it sure seems like there’s something else going on. My 2nd favorite whipping post is the 3rd wave style decoupling from reality trend. The more we rely on others to go out and interact with the world … the more we buy into their ontology instead of touching the world ourselves, the more likely we are to cherry-pick any old narrative that seems to line up with whatever bubble we live in. This doesn’t contradict the idea that it’s been going on since we grew a neocortex. If your shaman tells you to drink the muck, you drink the muck, whatever story she tells you about what the muck is or does. But IDK, hearing that anti-postmodern, group-think accusation from right-wingers and right-wing enablers cooonnnnssstttaaaannnttttlllyyy makes me want to avoid using it myself.

cultural niche expansion and comparative selection rates

No, Animals Do Not Have Genders [⛤]

The claim made in this article seems to be basically that cultural evolution is too fast to impact biological evolution. But my question is about how rates of evolution impact side effects and modularity. Why can’t we consider the idea that even though cultural evolution is fast, there might be long-term, stable patterns in culture that impact niches explorable by biological species? I.e. while the base rate of some dynamic might be fast, a higher order pattern (analogous to frequency beats in music) might be much much slower. And the rate of that higher order pattern might be near the scale of the base rate of some other dynamic.

The analog I have in mind is the gut biome. The base evolution rate of the bugs in your secum will be much faster than the base evolution of mammals. But, surely … Shirley!, the existence and general behavior of the collective microbiomes in mammalian guts has impacted the evolution of mammals.

Couldn’t, for example, the tendency for gender to arise mostly in binary form be a long-term, higher-order pattern that could impact the evolution of humans? Did early modern humans exhibit cultural evolution? Gender? If so, is 2-3 million years enough time for biological evolution to have changed us? Forgetting what happened before homo sapiens, if we assume 20k years for a significant mutation, 2m/20k gives about 100 possible changes. If early h.s. had culture with patterns stable for 20k years, then it would be possible for culturally defined niches to impact biological evolution. Right? Are my assumptions too simple? Is it possible but simply hasn’t happened?

More fundamentally, arguments like the following seem to affirm [⛧] the idea of downward causation:

Gendered Division Of Labor Gave Modern Humans Advantage Over Neanderthals

that gender may have helped another species go extinct. The sex-based, cultural phenomenon of gender may have changed the environment into which we’re relaxing in a significant way. Maybe the details make their argument wrong. But could there be a form in which it’s right?

[⛤] This CES module seems to disagree. But the discussion I’m trying to lay out should be independent of whether animals exhibit culture.

[⛧] Affirmation being much weaker than contradiction, of course.

OMG not God Again

I’ve reason to revisit my rejection of the term “atheist” as covering my own beliefs. This revisit is triggered largely by this video: becoming fools | a visit from some JWs [cc], wherein he talks about the JWs having a disarming naivety. My usual arguments for rejecting the term “atheist” as a descriptor of my beliefs is because I find uses for the concepts of gods and the various beliefs involved. My approach to the world centers largely in usage. And my understanding of the world relies fundamentally on privileging action over thoughts, one’s sensorimotor control surface over one’s inner life. Hence, if one uses supernatural concepts like gods or the Tao, or even shows preference for one metaphysical conception over another (e.g. many worlds vs objective reduction), then one cannot be “without god”, by some (several alternate) definition(s) of “god”.

Over the decades, I’ve tried to explain this to wide swaths of people types, from ardent believers (in one thing or another) to fellow atheists. I almost always fail. Lately I’ve landed on the qualifier “methodological”. A few posts and several years ago, I began calling myself a methodological atheist. And in everyday conversation, that works well because I rarely use concepts from the religions of the people I commonly run into (Christians, Muslims, & Jews — in that order, oddly). Their conceptions of god are simply not useful. I do find use for pantheistic, taoistic, Greek/Roman, Wiccan, and Lovecraftian conceptions, though. And it’s important that I use those concepts frequently to explain what I think about a lot of topics. So, I’m starting to lean away from “methodological atheist” as a good descriptor for me. As the surrounding community matures into Nones and Apatheism, I’m finding more and more reason to talk about these transcendentalist ways of being.

Now, after that overly long back-story, I find it interesting when someone’s authentic, dorky affect disarms a would-be critic. I think it’s a result of the asymmetry between discussants. Its most profound effects come in the form of Dunning-Kruger, with the 2-fold result that competent people over-estimate the competence of their colleagues and incompetent people over-estimate their own competence. But another pattern I notice a lot is the exchange between authentic liberal and conservative discussants. By “authentic”, I intend to talk about people who actually want to discuss ideas, as opposed to treating a discussion like a debate or a zero-sum competition. It seems to me that liberals tend to be more agile in their belief sets, whereas conservatives tend to rely on the structure of their belief sets. I.e. there’s more flex and slop (in the good sense of adaptive wiggle room) in liberal thinking. For a well-tuned, cognitively powerful person, this implies that conservatives (once tuned to a domain) may well be very efficient at solving problems in that domain. And liberals may well be very effective at switching domains.

In my experience, this difference leads to the liberals being more accommodating of alternative conceptions than conservatives. I.e. an authentic conservative’s rhetoric — or their naive belief in their own rhetoric — can easily disarm an authentic liberal. By contrast, I think it’s rare to find a conservative talk of being disarmed by liberal rhetoric. (They do talk of disarming affect and other things, just not of rhetoric.) Of course, if a tactic like that is adopted purposefully, then we have to demote them and strip them of their “authentic” rank.

To round out this post, although I shouldn’t have to say this, the mode in which I describe myself as “not atheist”, “methodological atheist”, “mystical atheist”, “agnostic”, etc. depends on my estimation of the authenticity of whoever’s involved in the discussion. And I’m more frequently finding that others’ authenticity is not easy to estimate. So I may well just “buck up” and align with my fellow atheists without the qualifiers, because anyone still arguing for any kind of theism just seems inauthentic, lately.

Leaving Portland

For the past decade or so, I’ve used Powell’s Wishlist to keep track of the books I’ve wanted to buy or check out. I did this because I like to think I’m committed to supporting local businesses, in spite of their website being painfully slow, along with other problems. But now I’ve left Portland. I will not rely on Amazon, because they are the antithesis of a local business; and I only use them when necessary. This leaves me in a bit of a pickle because even though Powell’s site sucks, it did work. Long ago, I signed up for Goodreads, but haven’t used it much. I suppose I’ll start using it, now.

But the point of this post is to wonder if there are any similar sites for journal articles (or news, or oddball garbage anywhere on the internet). I know there are sites for keeping track of academic citations. But I’m not an academic or a scholar of any kind. I just want to keep track of things I may want to read or watch someday. I have been using Reddit. But Reddit is full of snarky jerks and seems to encourage the race to mediocrity characteristic of all “social media”. I haven’t explored tools like Pocket, which Firefox was nice enough to introduce. I don’t know … I suppose I’m getting old. I’m simply lacking the energy to poke around a bunch of tools to see which one’s best for whatever given task.

Re: Why Books Don’t Work

In my schizotypal exploration of the graph that extends out from any given paper I try to read, I always pine for a less expensive way to learn all that … stuff, only well enough to read the original paper, of course. I criticize my friends whose first instinct is to “take a class” when they want to learn something, mostly because I’ve never learned anything from any of the classes I’ve ever taken … equivocating on “learn”, of course. Similarly, I see my friends whip through, say, Consilience by EO Wilson, which took me 3 times as long, and as much more effort, to read. And I think “What’s wrong with me?” An interesting side-effect is that most of the books I’ve read have my scribblings (in pencil! … I’m not a monster.) in the margins and any blank space in the back offering naive objections to silly tangents and brain farts I had while reading it.

It’s gotten so bad, now, that I hardly even read books anymore. If/when I do, I read like 10 books at a time, hopping from one to the other, conflating whose idea went where and assembling a Rube Goldberg machine conception of whatever the hell it was I was exploring.

In any case, I purposefully decided about a year ago to spend more time playing video games. One of the interesting features of my game-playing is that I much prefer playing the same game several times over serially playing different games. There’s something about the way these modern games are designed that helps them “unfold” as you play, or alternatively allows you to “relax into them” as you play. It’s very distinct from the “read the instructions then play the game” approach I was used to as a kid. It’s true that I prefer less linear, more open world games. But even so, it seems fundamentally similar to how I learn any other thing. There also seems to be some “aliveness” aspect. Even if I tend to play alone, multiplayer games are way more interesting … more learnable than single player games, which simply feel dead, no matter how well-programmed their AIs are.

The article cited at the top touches on all this nicely. But there’s still something missing from his conception. And I can’t quite put my finger on it. It has something to do with the dynamism of knowledge. Static knowledge isn’t knowledge. It’s somehow impoverished … maybe call that dead stuff “information” or “data”. Knowledge is somehow “alive”, explorable, open-ended, and non-linear.

On the noncommutativity of Social[ist] & Democra[t|cy]

I thought that I’d written about this before. But maybe it was elsewhere. In any case, this article turned me on to the noncommutativity of the two concepts. Further, this paper helped me clarify why I’m more inclined to qualify democracy with socialism, rather than qualifying socialism with democracy.

While Liu contrasts democratic socialism against state/authoritarian socialism, he doesn’t go far enough toward Gintis’ implication that democracy is (what I’ll call) a government’s interaction with reality. Gintis’ claim is more specific, that democracy is how the polity holds the politician accountable. But as we’ve seen with Obama’s detritus (well identified by Liu) and the election of Trump, the relationship between the polity’s plaintive noise and what happens to the politicians that fail them is not at all clear or direct. Obama won’t be held accountable for his wimpy, COMPROMISED attempts. And while you might say the Democratic Party was held accountable, I still don’t buy it, even with vagaries added.

But I do believe that democracy is what allowed our polity to moan and complain loud enough, and vote a weirdo incompetent Trump into power. The election of Trump clearly demonstrates the disconnect between our previous politicians and the reality they’re attempting to govern. How/if/what they learn from their new knowledge of how decoupled they are is an open question.

But back to the point. I believe any fossilized socialism, whether it’s fossilized into laws, or bureaucracy, or party in power, or whatever, will always tend towards a decoupling from the reality it tries to govern. Liu relies on some democratic spice, sprinkled on top of the socialism to keep it in touch with reality. But I think we’ll see analogous failures in democratic socialism (DS) that we’ve seen in social democracy (SD). If the capitalists retain their elite status in SD, then the statists will retain their elite status in DS.

And, for myself, having (recently) matured out of neoliberalism, I’m much more comfortable with the noisy and chaotic (perhaps Schumpeterian) ecology of VCs, indie programmers, banksters, “entrepreneurs”, the gig economy, and multi-national corporations than I am with relying on fossilized infrastructure with a peppering of democratic spice.

Nothing to say

When I was about 15 years old, my dad (yet again) reiterated his belief that I had no people skills and needed to work on them.  Both my and a friend’s parents used to joke that they didn’t know if/when our voices deepened because they never heard us speak.  I stayed in that mode until about my 2nd year in college, when I finally found some things to say, mostly to the Christian wackos that regularly manned a foldable desk in the student union.  From that point forward, I began to act on my dad’s advice and experimented with my interactions with others, exploring the gamut between complete silence and gregarious (angry or happy) expression.  I put those social skills to good use in my roles at Lockheed Martin and the Santa Fe Institute.  I used them a bit at the dot-coms I worked at later.  But I’d begun to lose my enthusiasm for it.  This and previous web logs and email lists were/are the last vestige of that meager desire to express my opinions.

I now find myself with a nearly complete lack of such desire.  Nothing to say.  ∴ Nothing posted.