madeofmeat
No bio...
User ID: 1063
I feel like there's something tricky here. There used to be the thing where people were going "schools should teach critical thinking", that certainly sounds like a cross-domain understanding of sorts. People actually tried to do this, and it turned out that it's either very hard or impossible with the existing toolkit of teaching domain-specific stuff. Maybe it can't really be taught and some people just pick it up by themselves, maybe it needs one-on-one tutoring that doesn't scale.
It's also tricky to apply a fuzzy "might be relevant to business success" / "probably isn't" judgment to rigid curricula and socially recognized pursuits. People will want to legibilize things into clear-lined singular pursuits like "playing tennis" or "being an accountant".
The success stories for very early hyperspecialization seem to be very "inside the box" things like playing the violin or being good at golf or chess. You know exactly what you're supposed to do, what is and isn't allowed is tightly circumscribed, and mastery generally just involves knowing as much stuff inside the allowed box and being very well trained at executing it. Things like business or science aren't like this. You are allowed to come up with completely new things for both what you're trying to accomplish and for how you're going to do it. Arguably there's still a box of physical reality and the laws of nature, but those aren't exactly easy to start getting a hang of at age five, unlike "what are the rules of chess" or "how do you hold a violin". If you want to do the sort of cross-cutting paradigm-busting that pushes things ahead, having been hyper-specialized into one of what your parents' generation thought was the set of relevant schemas for succeeding in the world might not be that helpful.
The mods include the site administrator, so it's kind of a given they can even if it weren't stated explicitly. Sharecropping sites like Reddit where the mods are just slightly more privileged users and the people running the site aren't cooperative are a special case, not how most internet sites works.
Maybe don't try to force yourself to be "widely cultured", but lean in on specific interests hyperfocus instead. Try to find a thing or three that are not stereotypically low status nerd culture, but also obscure enough that you're not likely to run into anyone else being into this specific thing. Like medieval Chinese painting, Roman poetry or political theology in the Byzantine Empire. Poke around anything older than 50-100 years and then when something looks interesting, just dive all in on the rabbithole of that specific thing. The plan is to come off as more of a foreigner of the same social class, you're not quite versed in the same stuff everyone else is but still giving the roughly correct vibe, rather than an easily pigeonholeable weeb pleb. If you can find some specific thing with good cultural valences you can get yourself to be genuinely interested in, that's going to be a huge force multiplier with actually getting deep enough in the thing for it to do some good.
(I think Zorba should fix the system so that AAQCs don't get flagged the same as "Reported").
The potential Quality Contributions are the best part of janitor duty.
It's good for some things that can be solved with ten lines of code and that only depend on commonly understood concepts like files, databases and dates and not, for example, the architectural details of a private codebase. How complex of a game are you thinking and how do you see yourself describing what you want in the game to the AI? Games have lots of trickiness going on with them, they grow up to have idiosyncratic codebases where you need to understand the local architecture, bigger ones are logically very complex and all parts need to keep making sense given the overarching design, they're by definition somewhat unique and you often need to do many rounds of iteration to get the code to correspond to the behavior you want them to have.
The link to the "I don't see any value in the HBD hypothesis." comment is broken by the ?context
parameter (because of deleted comments upstream?). This seems to be a working link.
In general, I wish TheMotte's comment linking would work more like reddit's or Hacker News' instead of trying to force the context-parameters and #context
everywhere (what's this even for?). Support a permalink that's a post-url/12345
(add a 'permalink' link under the comment that gets you this) that shows only the 12345
post and it's children and an anchor link post-url#12345
(you could make the timestamp of the post clickable and give you this) that shows the whole thread but centers on the given post. Don't add extra "show context" parameters unless the user asks for it.
It makes sense that if the mod started out as a regular participant in the conversation, they should be hesitant to switch to modhat posting. When the first thing the mod posts in the conversation is a modhat post, it doesn't make sense that they'd need a second mod to make more modhat posts.
The other problem is that even if the "because genetics" explanation brings compelling evidence (which definitely happens) it isn't actionable except to refute a "because racism" explanation that was already lacking supporting evidence.
It suggests an actionable solution of researching gene therapies that increase intelligence.
Also, setting up some sort of baseline welfare state and somewhat paternalistic social institutions instead of engineering society with the assumption that everybody could train themselves to perform a well-paid knowledge work job and consistently make rational personal decisions if it weren't for moral failings like laziness, and that the people who don't manage that deserve what's coming to them.
The closest I've come to encountering a coherent proposal from "group average aficionados" is on immigration policy, generally taking the form of blanket/severe prohibitions against immigrants from countries with low average IQ (or whatever). But if IQ is of such vital importance, why not just test for it directly rather than relying on a crude circuitous heuristic? I took an IQ test myself and scored extremely high,[4] so what do you gain by overlooking that in favor of the purported average of ~37 million people?
I don't see why you present this part as a big gotcha. My first instinct is to say "that sounds great, let's do exactly that!" Bit of a problem with further thought though is that IQ tests mostly work because they're currently low stakes and there isn't much incentive to try to get good at gaming them. If you suddenly made them a pivotal load-bearing component on a very important and desirable thing, you'd get an overnight IQ test prep industry popping up, with all the existing tests immediately leaked to serve as practice material. You'd still get some signal, but I'm pretty sure months of practice are going to skew IQ test results. I'm probably still on the side of trying this, do it for a while and see how much of a problem the test prep ends up being.
Progressives are already viciously allergic to accepting the conclusions that naturally flow from their own worldview.
This last part feels like it takes a bit of a swerve with the argument and I'm not sure I see how it fits in any total thesis for the post. It feels like it maybe should've been a whole second post. Looks like you're gesturing towards a wider pattern, I guess seen in The Cult of Smart too, that depressed IQ is gonna depressed IQ, even if it's environmentally caused, with all the expected bad effects for life outcomes, but progressives are basically just equivocating accepting this into full acceptance of immutable hereditary IQ differences and denying both with equal vehemence. It's certainly a possible angle of attack, but it seems that if you want to keep talk of the possible genetic group differences off the table, we'd still be mostly in the status quo where people will just aggressively go for the "genetic group differences are impossible, actually" angle, since arguing back against this is not allowed. They can then just go back to playing the endless game of claiming structural racism and use the noise from this to draw attention away from practical problems like what you pointed out.
Ping @ZorbaTHut, the glossary looks like a valuable project, is it possible to raise the post length limit for this specific post?
Okay, how's this:
The belief that genetics cause significant individual differences in socially significant mental traits of people, such as temperament and intelligence, and these differences may be difficult or impossible to change with environmental interventions. For example, people might have an innate level of intelligence that cannot be meaningfully raised and that isn't high enough for effectively learning complex and high-status jobs for many people. Opposed to a belief held implicitly in much of mid-to-late 20th century sociology and cultural anthropology that differences in such features are culturally determined and can be fully remedied with environmental interventions like extending compulsory education and policing racist microaggressions. The thing Steven Pinker writes about in his book The Blank Slate and Charles Murray in Human Diversity.
Suggested addition: The various [color] tribes, red, blue and grey, and possibly violet.
Ingroup / outgroup / fargroup for that matter, and the weird dynamics that follow, like blue tribe turning a blind eye to Islam because they see it as the irrelevant fargroup and red tribe as the existential risk outgroup.
The HBD entry struck me as a bit weird as well. It doesn't feel like it's difficult to define, like something like 'Moloch' might be (could also be added to the glossary btw), so it's confusing why you seem to be playing coy with this one thing all of a sudden.
Also, the "von" in von Neumann should be lower case.
I think what's currently being done is that first someone does a GWAS study to figure out the markers for whatever we want to control, like height. Then in vitro fertilization techniques are used to make dozens of fertilized eggs, they're all sequenced, and you look at the markers given by the GWAS study on each and pick the one you prefer. That's not do-everything genetic engineering yet, you're still rolling the same old dice, but now you can roll them on the lab bench, look at some of the numbers early on and pick the one you like best.
I thought people were reacting to the woke reverse discrimination policies and systemic racism discourse that are driven by the argument that because "blank slatism" is obviously true, we need to be obsessively following societal outcomes by race and interpret any disparity as necessary proof of racism. Charles Murray says this is the reason he wrote Facing Reality, the guy Hanania is chiding claims that the woke "equality thesis" leading to unending recriminations about white racism as long as outcome disparity remains is why he's writing his stuff.
Most people seem like they would be happy to go back to a 1990s style color blindness detente, assuming it was applied evenly. You don't talk about race and intelligence publicly and you don't do racial grievance identity politics. What we got instead is that people took "there are no racial differences in intelligence" as the implicit uncontroversial truth, and drew up the whole intellectual edifice of systemic racism from that. What this has already cashed out in practice has been a complete travesty of "evaluating people on the basis of individual ability/merit", and this shows no sign of stopping. People just go "what can you do, we must keep fighting against systemic racism until it goes away".
The HBD discourse is fueled by people freaking out over representation in high-status classes like successful CEOs, scientists and Ivy League applicants. People get into those by being outliers on a standard distribution in talent, and this amplifies small initial population differences. A relatively small difference in mean IQ between populations that looks inconsequential in everyday tasks can still end up making the tail of people capable of becoming theoretical physicists have a 2x overrepresentation of people from the higher mean IQ population.
Greg Egan's Permutation City is a classic example. It starts out with your usual Matrix-like virtual world thing but then gets weirder than that,
That’s my whole point. I suspect that there is perhaps no sequence of events that could happen that would convince the Science crowd that NHI exists.
A Russian family that had cut all contacts to civilization and lived in the woods as hermits figured out Sputnik from seeing a new "star" that moved very fast across the sky. There's a lot of very observable new things you can make happen if you want and have the technology for it.
What qualifies someone to be a major public intellectual?
People from outside their religion regularly taking their arguments seriously would be a good start. Modern Christian apologetics don't seem to be having much headway with people who aren't already looking to be sold on Christianity.
Who's Harding?
If I hadn’t read too much Hanson over the years I would likely be a true believer in the alien stuff.
What do you make of Hanson apparently thinking the alien stuff is legit nowadays? As far as I understood, he thinks "they're here but we don't see them colonizing the galaxy" can make sense if there was just one abiogenesis event in the Milky Way, it led to both life on Earth and an alien civilization somewhere reasonably close in the stellar neighborhood through some sort of panspermia process. This would mean the single alien civilization that's contacting us has no serious competition driving it to visibly messing up the galaxy and is close enough they can reach us with less impressive interstellar technology that one that turns an entire galaxy into Dyson spheres.
- Prev
- Next
Ever tried reading old philosophy? Plenty of relatable cantankerous depressives like Schopenhauer overthinking stuff out there, but possibly ending up with some genuinely interesting viewpoints instead of just stuff you could fill in yourself starting from "guy's depressed".
More options
Context Copy link