GaBeRockKing
No bio...
User ID: 3255
He should have refused to certify the election
Unless you actually and truly believe the election was stolen, and you can prove it (to the satisfaction of the voters, not the courts), that would have been a preposterously stupid plan. Remember: democracy is just a proxy for civil war. Parties prefer to use the proxy when things are close, even if they lose despite in theory posessing a military advantage. (See: republicans not declaring war despite the fact that they control a majority of fighting-age men, and democrats not declaring war despite controlling a majority of the total population.) But the instant you click that "defect" button, your opponents do too. Unless your evidence is so convincing the majority of the other party will tuck their tails between their legs and admit malfeasance, refusing to certify an election is exactly equivalent to starting a civil war. Which would be a truly stupid thing to do, unless you control such a proportion of the population that you can expect total victory.
Equivocating autocratic control over one of the most potent mass-media apparatuses ever creating with "mean tweets" is disingenuous and you know it. I won't pretend leftists care for any high-minded free-speech related reasons, but frankly it's perfectly reasonable to fear and despite anyone with the kind of power elon musk has regardless of their ideology.
Every politician should be a mix between "leader" and "representative." It's up to the voters to choose exactly how they want their mix tuned.
Trump's populism has recently proven the current electoral effectiveness of being representative. Democrats have adapted to the meta.
I think gender differences have more to do with which groups each party and politician is better at representing than any preference for different styles. You could use gender essentialist framing to argue that women should prefer a "leader" because they're less able to lead themselves and be equivalently wrong because either way, it's a just-so story.
Late response, but...
The world's cultures are massively heterogenous at any specific point in time, and 5000 years of unprecedentedly rapid social change only increases the heterogenity. What "worked" for the past 5000 years wasn't any fixed culture, but a massive variety of intermeshed cultural niches developing in both cyclical and progressive ways. Even trying to reason from some sort of hypothetical majority culture is fallacious. Some adaptations become useful precisely because they're in a minority. Raiding farmers from horseback stops working if you kill all the farmers, but that doesn't mean you should become a farmer yourself.
"Norm-shattering" is a good description, but incomplete. What trump is is the first person to recognize and reify a wholly new strategy for executing politics. That being: to organize his supporters not as interest groups, not as a cult of personality, not as ideological compatriots, but as a fandom.
Recently, we've been seeing a lot of ink spilt on the subject of the social media-depression link. Particularly where it concerns children, but I hold that the problem extends universally across age groups. Ubiquitous smartphones with social media is (so far) the ultimate realization of the "bowling alone" trend-- where the world inside the screen becomes so addictive that people lose social links outside the screen. Consequently, in-person social links become scarce despite being just as valuable as ever. More valuable, perhaps, because in-person interactions retain all their old benefits while also making you a high-priority person to someone who has potentially valuable virtual contacts.
People on some level realize this, so they still optimize for some level of in-person contact. And they do that by engaging in fandoms. Large, energized masses of people with a shared understanding of a universe easily gel together when they meet in person. That fact that these universes are fictional doesn't matter. In fact, the very fictionality of these universes is what makes them so effective. They can optimize for being interesting and pleasurable over being true. (See: epistemic minor leagues). And unlike traditional social groups that performed the same function (e.g., fraternal societies, religions) they demand very little from you personally outside what you were already willing to give: the free time you already wanted to spend doing something fun, and the opportunity cost of spending time with people who aren't into the same things you are anyways.
Trump is the first modern politician to truly realize the power of fandoms. I want to say, "unwittingly" because I think he's an idiot, but given his success with TV and branded enterprises I can't rule out genuine epiphany. He's creating a shared universe than his fans can all be passionate about, with interesting characters, noble heroes, and evil villains. And in organizing his political supporters into a fandom, he's invalidated all the usual tools of traditional politics. Fact checkers; negative news coverage; research papers-- none of that stuff is effective against a fandom. In fact, it's actively counterproductive. Every youtube video about how star wars physics aren't realistic just keeps people interested in the star wars fandom.
Indeed, the only thing that can successfully oppose a fandom is an equal and opposite hatedom. Whether his enemies deliberately organized themselves into one, or were simply forced by selection pressures to fit the mold doesn't matter. What matters is that when someone says, "Drake and Josh are amazing singers," you don't bother telling them that they're overproduced corporate slop. Instead, you go out and create a powerful social group of your own, by telling them, "look at all these idiots that love something Dan Schneider, a pedophile, created!"
The hate against the clintons for being slimy, the "bush is stupid" people, and the the obama birtherism were essentially prefigurement for this. They cultivated proto-fandoms with their charisma that made traditional policy attacks less effective, and therefore were subject to proto-hatedoms. Bernie came the closest to emulating trump with his own dedicated online fandom, but he definitely didn't consciously understand what was happening, and in any case failed to take advantage of his devotees like trump did.
Until technology dramatically changes the social environment again, I predict that every future president will act like trump and be treated like trump. He is to social media as Kennedy was to television as Coolidge was to radio-- laying out the path for every candidate after.
You don't want large swaths of states (large the rural parts) to be grossly less developed than the rest of the state. Culture and politics aside, this eventually results in economic degeneracy.
Speak for yourself. Subsidizing inefficient sectors of the economy for political reasons is ALREADY economic degeneracy.
If you're a member of category A and want to become a member of category B, the individual judgement you'll receive will be the product of how people feel about the rights and responsibilities conferred about groups A and B and their assessment of your relative fitness to bear those rights and relative suitability to execute those responsibilities.
So I don't think it's necessarily inconsistent for people to support transgenderism and not transracialism. I can easily imagine, say, a racist that believes in gender egalitarianism. Or in reverse, a racial egalitarian that believes in traditional gender roles. Or even just someone who isn't making ideological but tactical decisions-- believing in principle that neither transgenderism or transracialism matter because in the utopian future no one will see race or sex, but in the present advocating for differential treatment in some areas and egalitarian treatment in others as pragmatic steps toward achieving said utopian future.
One trillion would be a dramatic underestimate. Being the smartest person out of the nearest million credibly puts you in millionaire-tier even before network effects. Geniuses competing with geniuses would make massive strides in every concievable theoretical field before the decade is out.
Plus, most societies have to deal with the fundamental constraint that norms have to be designed so that unusually smart people and unusually dumb people can't screw things up for everyone else, which adds a lot of frictional inefficiency to every interaction.
What I am disagreeing with you is the idea that this will lead to some kind of average IQ convergence- more than likely the black/white gap will grow for a bit and then stay about the same.
To the extent that any IQ difference is caused by genetics, we should expect to see that IQ difference decline as greater admixture rates have been achieved. To the extent that IQ differences are caused by environmental factors, we should expect to see that IQ difference decline as those environmental differences have been declining.
For the gap to stay the same or increase despite changes to the putative causative factors would imply that something is off about our understanding of IQ and/or racial IQ, though I couldn't say in advance what.
Including my responses to your other comment here
I mean, the hybrid vigor hypothesis for IQ specifically seems pre-falsified- most of history's greatest geniuses were purebred members of endogamous groups
There’s a large number of countries where mixed-race people are the majority(most of them Spanish speaking) and the best and brightest from these countries seem to, generally, be not mixed, often specifically descended from high-IQ immigrants.
Re-reading my original comment, I think you're right that my hot take doesn't follow from its premises. So I'll alter it a little. "If theories about racial intelligence are correct* we should expect..."
(Here taking "theories about racial intelligence" to mean specifically theories about a genes that have reached fixation due to selection pressures at the race level.)
So far, you're right that we haven't actually seen much evidence of hybrid vigor in the domain of intelligence-- which is kind of the problem with these theories about racial intelligence. If you told me, "different corn breeds have different disease resistance capabilities," then I could posit that "it should therefore be possible to engineer hybrid varieties with superior disease resistance to any heritage line," and prove us both empirically correct. But I'm hearing "different races have different intellectual capabilities," and yet not seeing any of the superlative hybrid strains. I'm aware that isn't proof for the negative case (that no genetic intellectual difference exists), but it makes me unwilling to reject the null hypothesis.
If you (or the original study) do in fact convince me of racial differences, I'll switch to believing that hybrid vigour should occur, and that we just haven't tested the right crossbreeds yet.
If we're talking about political economy, The People also hate hate hate to hear about innocent (or at least, insufficiently proven-to-be guilty) people being executed. They're also mad about people being falsely imprisoned, but to a far lesser degree.) There's no objective reason we should be spending our resources to satisfy the vengeance-lovers over the mercy-lovers, and plenty of practical reasons why we wouldn't want to encourage "vengeance" as a core value of our society.
Okay, I've found a graph from 2019 that compared against TFR: https://www.statista.com/statistics/1238575/total-fertility-rate-us-education/. There's also this graph that subdivides by race that shows different race-based curves, which is an objection someone brought up elsewhere.
The hook-shaped curve is evidence for what I'm talking about-- that selection pressures have returned to favor educational attainment (and therefore, by proxy, IQ). Especially since the true strength of the relationship is going to be obscured by older women with less education who already had all their children, and obscured by an inability to differentiate between then-young mothers who won't and will later get additional education.
Elsewhere someone brought up the objection that even if positive selection pressure in favor of education has begun to apply to whites and hispanics, it doesn't seem to apply to blacks, which would mean we should expect to see racial iq differences (if they exist) to continue to diverge. But I reject that reasoning on the basis due to the fact that base rates of education have always been lower for blacks for structural reasons (i.e., racism, poverty), that black TFR started from a higher modern basepoint and has dropped faster, and that therefore the existing obscuring effects of having an older, less-educated cohort would be stronger.
In the interest of honesty I also found this graph for birth rates in 2020 https://www.statista.com/statistics/195970/number-of-births-by-educational-attainment-of-mother-in-the-united-states/ but I'm having a hard time comparing it against the original graph because it splits things up into different brackets.
I don't think hybrid vigor would be a factor in every case, but alleles that manage to reach fixation in a particular population group likely make some sort of balanced tradeoff-- not optimizing for any one thing too hard because doing so would confer metabolic costs while producing no competitive advantages (against a population with the same genes.) So if racial IQ science is real* I'd suspect that in the right-tail of IQ distribution we'd see people with heterozygous alleles that were "intended" to perform some balanced function in isolated groups, but combined by crossbreeding lose any limiting factors and produce people exceptional in some way or another.
*To avoid concern trolling, I should be clear here that I do not have any particular faith in the existing racial IQ science. I can't rule out in principle that racial differences between IQ groups exist, but there are too many confounding factors, and in particular people who talk about race-based IQ support their claims with evidence that matches the pop-sci view of what "evolution" looks like, rather than what any plant or animal breeder would say actually happens in practice with genetics.
I think you're right in that there's some fuzziness involved in determining the effect size, especially since the correlation between IQ and educational attainment is strong but not incredibly so. But the objection you bring up means that the graph I posted should be underestimating the size of the new more-smarts-means-more-babies effect, since in the past fertility rates were higher.
At any given point in time and cultural context, a genetic trait might be positive or negative for fertility. But over the long run, everything reverts to the mean. If IQ is real, it should have some sort of consistent, visible effect visible over long time periods-- and 20 years is a fairly long time period. The reason I pointed to the graph of people with more educational attainment regaining their advantage in fertility is to display exactly this phenomenon. Truly dysgenic phenomenon are intrinsically self-defeating.
And it's worth considering that fertility rates based on genetic traits are a lagging indicator. That we're seeing evidence of a switch to benefiting probably-smart people means that the actual mechanism must have been in place for at least 20-40 years, since all the newly fertile PHDs must have had the genetic components of their IQs determined then.
The fertility rate inversion we're seeing is also not just an american thing, but common cross culturally as more countries go through the stages of the demographic transition.
All that is to say, I trust Darwin and Rudolf Clausius way more than I trust anyone else. Any scientific paper has to get through them before it gets to me.
Unrelated hot take: while mixed race people, as a group, will continue to be average, we should expect the leading geniuses of the mid 20th century to be disproportionately mixed race due to the effects of hybrid vigor. And if theories about genetic intellectual differences between races are correct, we should be able to identify specific combinations likely to produce particularly intelligent hybrids.
That's dumb. I guarantee your local morons running for city council aren't important enough to be puppeted by whatever evil group you think controls america. Yes, choosing between the soccer mom who thinks harry potter is satanic and should be banned from every school in the district versus the strung-out ex-hippy that wants the police to raise sales taxes by 0.5% to fund their vision of renovating the playground in park fuckhill is less glamorous that voting for GOOD versus EVIL in the national elections. But mediated by the fact that local elections are often one or lost by only hundreds or even dozens of votes, the compounded effect of voting in your local elections dramatically outweight any possible impact you could have in national elections-- and that's even if the national elections were actually composed of the good party and the kicking-puppies party.
I'm not saying "plenty of" in terms of "proportion of the total population." I'm saying "plenty of" in terms of absolute numbers. I suspect many of those people would be happy for time off their sentences in return for working in the field they're trained for on behalf of the government.
Like, not to put too fine a point on it, but that's sort of exactly what we do with hackers already-- there's an existing pipeline from "black hat hacker" to "government spook."
And "worthy of imprisonment" crime is a measure biased towards people who commit crimes with an impact toward a few, specific individuals, rather than e.g. financial crimes that often have vastly more impact than your average armed robbery but result in far less jail time.
Are you talking about not voting for president or not voting at all? Because as the other person said, regardless of where you live and what ideology you favor, you should still be voting in local elections. The more local the election, the greater your voice-- and the more important the people you're voting for. A president has a lot less impact on your town's crime rate than the policies of your sheriff, mayor, and prosecutor.
I'd be interesting to see a follow-up too, since any result other than a regression to a mean would imply that either the initial study or our very concept of IQ is flawed. With greater racial interbreeding ratios and selection pressures in that have returned to favoring fertility rates in people with high educational attainment, if there is any actual of race on intelligence, it should be gradually reducing.
Our justice system is expensive because it's poorly designed. Or rather, because it wasn't designed-- because it's just a long pile-up of compromises with no guiding ethos. And yet, despite that, if we assume we're not going to redesign it, then imprisonment is still cheaper than the death penalty. If we assume we are going to redesign it, then why no redesign it so that criminals directly repay their contributions to society?
People demand that prisons be punitive while at the same time squeamish about the exact nature of punishment. Of course that leads to poor optimization for economic efficiency. We could get a lot more efficient use out of prisoners if we were a lot more judicious about exactly which rights we chose to violate, while at the same time not losing our heads if the same measures end up making prisoners happy. For example, encouraging moderate cocaine use but then predicating their supply on being productive and compliant.
(I'm not saying that specific intervention would solve our problem, just using it as an example of the sort of measure no one is even willing to consider.)
I'm also addressing your comment here:
The kind of labor that can be efficiently done by slaves is mostly done even more efficiently by machines these days, and prison guards cost a lot more than just hiring society's existing pool of the poorest
... with the above. Historically, slaves did plenty of complicated, specialized work that required a surprisingly high level of education. In rome,
Other vital services were provided by literate slaves who served as teachers, librarians, scribes, artists, entertainers – even doctors.
That in the modern day compelled labour is typically done by people with only the desire for and ability to compel uncomplicated work doesn't mean we'd have to stick to that paradigm. We imprison plenty of lawyers, hedge fund managers, accountants, scientists, etcetera. It shouldn't be impossible to convince them to do work that's on net beneficial to society even if we have to pay them with cash or reductions to their sentences.
The enslaved soldiers definitely would have still been out-group after being enslaved, and if anything more prone to massive violence. You're correct that executions were also common, but I don't think either of us have the data to talk about "standard practice" in this case. And yet, an argument that says, "hard labor cannot be more efficient than execution" required a preponderance of the evidence, while my position (that hard labor can be made to be more efficient) requires only a collection of positive examples, regardless of how representative they are of average behavior.
The whole theater of appeals and waiting is expensive. It should be more expeditive
Of course it's expensive-- society is will to accept a very low P[innocent of murder|executed for murder] rate. What false positive rate are you willing to accept?
Some medications downregulate sensitivity to particular neurotransmitters in the hopes that doing so causes your body to upregulate production of said neurotransmitters. Similarly, it makes sense to me to try creating buildings that make people feel alienated in an effort to push them towards more community-minded behavior. I can't make any concrete statements about how well it works... but nevertheless, I miss the old St. Paul.
I'm abbreviating for conciseness. Of course there's a lot of other factors that go into what false positive/negative rates we optimize our justice system to accept, like the risk of a justice system being seen as "soft" encouraging vigilantism and the degradation of state power.
Forced labour probably doesn't pay for itself
I'm almost absolutely positive it could. The carceral system-- and in particular the prison labor system-- is inefficient because it has misaligned incentives at every level due to a complete lack of consensus about its actual goals plus institutional inertia from times with totally different values. If we're talking about making dramatic reforms anyways (which would be required to significantly streamline the process of executing criminals) then we could orient things towards actually making the existence of prisoners net remunerative for society.
Basically. Hating powerful people that promote an ideology you don't like is common (and rational) cross-culturally. See also: republicans hating the soros brothers, reddit right-wingers hating Ellen Pao, everyone hating on Zuckerburg at various points for various reasons, etc.
More options
Context Copy link