curious_straight_ca
No bio...
User ID: 1845
curious_straight_ca
3hr ago
curious_straight_ca
3hr ago
The good reason to have it is maybe people will only post interesting links, like this or the kind of links in this. Or even just links saying 'DALL-E 3 was released' or a 'YIMBY roundup for Sep 2023'. I agree they, mostly, posted uninteresting links in the past, which is why I argue that the poisonous links should simply be banned based on their content. But it's a shame that we miss out on some interesting links and interesting discussion just because many of the posters 'can't behave'.
curious_straight_ca
16hr ago
The population at large thought fantastical the telegraph, cars, oil, artillery, fighter jets, electricity, nuclear bombs, computers, and neural nets, a century before they arrived. They still came, and clever people predicted them.
curious_straight_ca
19hr ago
I would not accept slavery and outlawing criticism of government.
But because those are just generally bad and probably won't stop the asteroid, not because unilateralism is bad, so I don't see what's wrong with the original premise of 'AI seems to eat other ethical concerns on a large scale'.
Same response as my comment above (yeah, padded effortpost spam is bad, maybe llm).
Taking the bait and responding to ... what I can gather from the actual content. Your two-pronged approach to spam is 1) detect it and delete it with algorithms and 2) go back to more decentralized platforms. 1 works, but existing platforms already do it and it's an arms race. 2 is dumb because you need centralized teams of competent people making antispam algorithms. Users shouldn't have to make their own bespoke spam filters, or determine the right filters to use in a decentralized way, terrible UX, this doesn't solve any problems that centralized systems don't already solve.
Was this a thought experiment in how lack of a BLR leads to padded effortposts? Probably. Each paragraph eschews detail or sources in favor of sweeping pronouncements of importance. It doesn't feel like GPT to me but i dunno, could be wrong. I'd actually love to read a post about random historical relationships new texts could shine new light on, but each sentence should be specific and relevant new information, not this.
edit: post author was 'dmz', seems to have deleted both of their posts.
My main gripe with the BLR was that a bunch of the posts were 'grayuniwave: new COVID report shows masks and paxlovid cause myocarditis' or 'WOKE college OPPRESSES innocent REPUBLICANS' and 'Joe Biden is promoting wokeness on steroids'. The third one is real. And some of the links that weren't like that weren't inspiring either. The good posts were, almost entirely, by the people you'd expect them to be from (the mods, dase, etc), or the links themselves were authors I already liked independently. (obviously the upvote counts didn't agree with me here)
https://old.reddit.com/r/TheMotte/comments/q1jdta/meta_on_short_posts_and_good_posts/
If I were the BDFL I'd just have a BLR and give various volunteers a mod power to specifically delete BLR posts they thought were 'bad' (and would code that feature). But that's quite against the stated (if not the actual) ethos of neutrality and tone not content.
Great article.
It seems unfortunate, in a grand sense, that so many rare books are sold without being scanned and uploaded to libgen and friends though. For every person who found exactly the book they're looking for at that shop, probably a five more didn't because they weren't there. And even if that'd undermine the financial self-interest that drives people to collect the books initially, there'd be no deadweight loss if the books were scanned but kept private for however many decades.
Which also brings to mind - there are many who find fulfillment in professions that are, in a fundamental sense, obsolete. It seems to me that this is, somehow, wrong - it's better to push at, and likely fail at, the frontier, instead of lingering in the dust, enjoying oneself. The former is tougher, and more interesting too, imo - and simultaneously materially benefits all more. But it seems likely that technology, AI, etc will consign us all to the latter in time.
curious_straight_ca
1d ago
Object-level: Obviously, Effective Altruism's concept of effective charities is what you want here. The usual place to start is GiveWell, but all of their top charities work mostly in Africa. Individual EAs often donate to other specific causes though, and there are many smaller projects and charities catalogued in various places. There's GWWC, maybe the places open philanthropy gives grants to, maybe just browse posts on the EA forum. Some of them do most of their work in specific places, some of which are in Asia, so you'd just look around there.
I think, even from a universalist hedonist utilitarian perspective, the longtermist idea that something like AI safety research or governance work is more important than malaria nets is very compelling. Marginally preventing a few dozen unnecessary deaths vs playing a part in shaping the entire future. Or at least, it would be if AI safety research and governance work was net positive or doing anything important, which isn't obvious. But however you approach it, the issue of AI and future technology transforming everything does seem to eat every other ethical concern if you think enough about it. Holden Karnofsky, previously co-CEO of Open Philanthropy, recently stepped down to focus on AI and is now "Director of AI Strategy".
curious_straight_ca
1d ago
Back on reddit, there used to be a 'bare-link repository' where you could post just links with minimal commentary. It was filled with low-quality and too culture-warry posts and the mods decided to drop it. I agree it should return, with heavier moderation of some sort.
Armenia is at least a real place, and Azeri a real people
The retort writes itself. The land Armenia sits on is real, sure, but the material importance it has is limited - drop the exact same population into Kansas or Nigeria and the ethnic Armenians would be able to farm and organize just as well. A communist who defends communism that happens to sit on a particular piece of land is different - how - from someone who holds an idea of a blood-and-soil tie to a particular piece of dirt that - in a literal sense - they'd be able to hold perfectly well on another piece of dirt? At best, the 'soil' is a metaphor for the people and their folkways. Unfortunately for Azerbaijan and Armenia, and every other country on the planet, material concerns have caused them to adopt the "folkways" of modern America and western Europe. Can you even name a concrete way of life or difference in genetic tendencies that differs between Armenians and Azerbaijanis? One that's worth killing over? And those are ... still differences in ideas. If Europeans are slightly more individualist than the Chinese, how is that more real than an entirely different political philosophy and strucure? And the difference in 'people' between Armenia and Azerbaijan are caused more by historical geopolitics between greater powers than they are anything else.
Say what you will about Ted K, he at least had coherent and falsifiable ideas. "Vitalism" on the e-right is nothing but a haze of grievance and nostalgia for aesthetics disconnected from their original application, and it breeds intellectual carelessness in its adherents that rivals right-populists and 'woke's.
but absolutely not for something as bloodless and fallible as “ideals“
Concrete peoples, nations, and homelands are exactly as fallible as 'ideas'. Is the Azeri who dies to annex an unimportant province of Armenia glorious? Or soldiers in the 300th post-colonial African coup? Is the individual soldier still glorious if military success is granted entirely due to alliances and imported technology and not the military prowess of their soldiers? Does it matter that these globally insignificant squabbles only serve to dig holes of economic instability and irrelevance deeper? War is itself a technology, and the material cause of its glorification was spreading genes and then growing states. Is it still the most effective means of doing either of those?
but I disagree the costs are worth it and also disagree with your cost estimates
I think that management that's competent and willing to innovate, combined with using new technology, would bring down costs and reduce side-effects a ton. People as smart and self-driven as those that've driven SV innovation for the past decade could come up with, try, and iterate on new solutions. I'm not sure what it'd look like - of course - but here's another idea: Mass deployment of improved versions of actually-currently-existing opioid vaccines, which have been under development for the past decade. "Antiopioid immunopharmacotherapies (e.g., conjugate vaccines) that sequester drug peripherally, preventing opioids from reaching targeted receptors in the brain". Current versions seem to last for around a year. Maybe offer free drugs at a clinic for a week if you take a long-acting version of that at the end. Maybe there'd just be an arms race between novel synthetic opioids and the vaccines (which the synthetics would win). But my main point is that it's easy to imagine a stasis where the government's options are to keep doing what it's done for the past six decades, which is a losing position, but it's possible for intelligent people to create new techniques and social systems that change the rule of the game.
And prisons are the opposite of 'competent management and willingness to innovate and experiment'. (effective innovation in coercive government will look pretty different than tech innovation though, it's a lot easier to write a thousand lines of code and deploy it than it is to create new physical infrastructure and train a few thousand people).
curious_straight_ca
5d ago
@ymeskhout curious if you have thoughts on my parent comment? Essentially, from first principles I really don't see why the 'war on drugs' is unwinnable, it'd just take sacrifices that are (imo) very obviously justified when compared to 100k overdose deaths per year. However, first principles reasoning goes wrong as often as it goes right, so if my speculation is obviously wrong I'd love to be pointed to something explaining why.
curious_straight_ca
5d ago
I also feel like shoplifting for resale is just not done here?
That might be an unstable equilibrium, though.
It's not politically viable today, but a much better solution is YIMBYism with well-enforced standards. Want to build 100 new apartments? Go for it. Want to openly screen applicants for looking clean and being the right social class, even if there's disparate impact? Want to kick someone out of one of the apartments for doing drugs, or imprison them for a single instance of petty theft? Go for it. To an extent NIMBYism is a response to communities not being able to more explicitly police their members (and some members not wanting to), and 'don't let anyone in' is a maximally illegible way to keep out the undesirables.
You are correct that it's wrong to not consider the trade-off and weigh both sides of it evenly. But - are there other potential trade-offs? In areas other than endless compassion and enabling?
You can calculate how much "drug use enabled due to clinics" is worth how much "unsupervised drug use and property crime". It's a reasonable question to ask. I think it's between plausible and likely that - provided someone more like Hoff than like a prison abolitionist administered the program, and tried to minimize growth of the drug user population - free drug clinics would be better than the status quo.
Here's another trade-off: How much loss of privacy, and increased incarceration, is worth how much heroin addiction? Say you can imprison one person for ten years to prevent five cases of (severe) addiction. Is that worth it? IMO, it is. And it's especially worth it when you realize that the benefit of law enforcement is almost entirely deterrence - deterrence of more organized and competent forms of crime that don't currently exist at all, but would naturally evolve without law enforcement. Things like deterring megacorporations from advertising and selling heroin on an industrial scale, or deterring gangs from capturing factories and imposing taxes on their output (as happened in post-soviet russia). I think it's entirely possible to move drug gangs and drug dealing at scale into that category. And after that's happened, harsh policing of drug gangs won't be any more necessary than harsh policing of other rare crime is, just monitoring for new small-scale enterprises and nipping the buds.
IMO the best policy is sacrificing significant privacy in high-drug-use communities, and freedom for those involved in distribution, in exchange for preventing most use of hard drugs. A hundred thousand overdose deaths per year is just ... a lot of death, and signals even more suffering. And when we directly compare the imprisonment and loss of freedoms/privacy to the cost of clinics - being a drug-zombie is still very bad, and being a managed-drug zombie in a clinic is somewhat less bad than being a street drug-zombie, but it's still terrible. And being imprisoned is probably 'as bad' as being a drug-zombie, and many fewer people are involved in distribution than are users. (Plus, if you really were able to entirely clean up distribution, you could quickly free the low-level distributors).
The principle here is that the state, when unencumbered by internal opposition and the details of the last century's procedural respect for rights (but stil lretaining the spirit), and using modern technology, is entirely able to prevent the sale and use of drugs at scale. I'm not sure what form it'd take, but I'm quite confident that computers + monopoly on force > decentralized drug networks if the state is willing to innovate and modify due process. Drug dealers are smart, but they don't have the power of mass surveillance, of spending X% of GDP legibly, of creating new social systems all members of society participate in, of targeted and overwhelming force.
This paragraph will be 'my vivid imagination' and 'obviously wrong and stupid' rather than a practical solution, as I'm not a subject-matter-expert and haven't spend 40 hours reading relevant material, but: We already know which areas have drug problems. To fix the problem, you need (roughly) two things: detailed knowledge of who's selling/distributing the drugs and where and when, and the political/procedural will to arrest and imprison them all. I'm assuming away the second issue (I don't think you need to abolish due process or anything). The first issue is still a problem, though - cracking down on the most open drug markets will just push the trade to less legible areas. So, in the 'bad areas', impose mandatory drug testing (fentanyl, not weed, and the tests that detect if you've used in the last few months) for everyone, every X months. Or maybe put drug metabolite monitors in everyone's toilets, or something. Those who test positive won't be punished at all. Probably, to make this incentive-compatible, an ideal state would pair this with temporary free drugs. But then use the high-resolution information on who uses and doesn't as a seed for using surveillance to identify low level distributors. That gives the enforcement arm enough information to just arrest everyone involved in distribution. Everyone involved gets put into newly built and very nice nordic-style prisons or something. You'd start by rolling this out in one community, see what works and what doesn't, iterate, and then make it larger-scale, of course. This would be expensive. But how expensive? Maybe 2% of GDP expensive, definitely not 10% of GDP expensive. We probably lose 1+% of GDP already to services, treatment, law enforcement, and lost productivity due to addiction, this link (which I do not trust at all) says 2%. And once most existing organizations were shut down, you could scale back all that spending and surveillance.
Obviously, that sucks for everyone imprisoned, and both the users and non-users who lose a significant amount of privacy. But does it really suck less than enabling many current addictions, and maybe minting new drug-zombies, because the free drugs clinic is just right there? I think if you just add up the utils, the loss in privacy and imprisonment of some distributors is more than made up for by the number of addicts getting divided by ten.
(An ideal state would also fix whatever the root causes of the massive wave of addiction are, it's not just drug presence. But the drugs play a large role, and there's significant reflexivity to it, most current addicts would stay addicted even if given the minimum level of welfare or life-purpose or w/e necessary to prevent them from getting addicted initially).
edited note: Also, I'm not claiming any of this is achievable by the current way the US government is structured, or that people should vote for policies similar to this today. Clearly, naive attempts to act on 'just arrest the drug dealers!! how hard can it be!!' have failed, even if the impulse there isn't entirely wrong.
curious_straight_ca
7d ago
You keep posting the same thing. The fact is, there are many happy couples with very below-average looks. The wife of an ugly man is not, on average, 'deeply disgusted' with him. (I'm not sure what effect this has on e.g. cheating, any observations will be very confounded by the association of unattractiveness with other things).
You're a decent writer, you seem capable of having interesting ideas. Do you have anything else you might be moved to write about? Why not try that? Maybe just vignettes from your life like george_e_hale, maybe some interesting technicality from your job, perhaps a commentary on ancient philosophy. Just anything else.
curious_straight_ca
7d ago
To avoid the kind of stagnation that tends to result from that, why shouldn't we adopt something like Georgism, which would weaken land-based property ownership within society but attempt to make it fair going forward?
The US economy is simply not stagnating. We have the highest nominal GDP/capita worldwide among large countries, lead novel research and industries (most recently AI), and our #1 competitor is notorious for stealing our IP. The US's property economy is stagnating for all the usual YIMBY reasons and a lot more construction should be legal, but that's not at all analogous to your OP. Georgism, taken literally, means replacing all taxes with a tax on land, which I don't think is workable because a lot of important capital is entirely intangible in the form of human capital, IP, and organizations and a land tax totally misses that.
curious_straight_ca
7d ago
It's only reprehensible because we have better alternatives, right? Say (tortured hypothetical) it's the year 1500, and you're picking crewmembers for your voyage in a week. You need a bunch, so all you have to evaluate them is a three minute chat, and they're not particularly educated so there's not much shared knowledge to go on. It's your experience from past voyages that jewish crewmembers are the best and africans are the worst, and this is still informative even after adjusting for your first impression. (Also, a bad crewmember might mean 'your ship sinks'). I think making the race-based judgement here is fine. The only other alternative is picking randomly to an extent. And if you have a moral issue with some people being deprived because they're paid less - how is 'not being chosen because you're black' morally worse than 'not being chosen because you're low iq'? Both are unchosen.
This, of course, isn't really true in the modern day. You can just give someone an IQ test or an interview problem or something.
The island collectively voted to transfer ten percent of the island's economic output to those who are less fortunate, funded by taxes on the more - and that doesn't include infrastructure, education, etc for the general benefit. And the original gang's descendants aren't in power anymore - they invited successive waves of immigrants, and, via free trade, transferred most of their property claims to the smartest and most industrious (or unscrupulous) of them. How many Jews, Asians, or Indians were on the island when the first nine people arrived, compared to how many are rich today?
Yeah, if we were still feudal or WASPs owned 90% of a rentier economy because they got here first, that'd be very unjust and inefficient. But that doesn't resemble the current state of America. A different argument against private ownership of the means of production - that it's exploitative or inefficient for those who come out on top of a fierce competition for profits to dictate much of the economy - at least accurately describes the thing it criticizes.
For people who defend the current conception of property in the industrialized world, and who think that we should accept the idea of starting at step 3 and not worrying about 1 and 2, what is the justification?
Because said free exchange has, several times over, redistributed all the capital from the original owners to those who won in a contest of merit (or underhandedness), and to whatever extent the original gang members hold more capital than some other groups, it's mostly because they have e.g. better genes or cultures - and, indeed, more recent arrivals with better genes/cultures have higher average income/wealth than the founding ethnicities.
A funny modern development is gambling streamers. Twitch, the incumbent in streaming, banned gambling a while ago, but it draws enough users to gambling that Stake, an online gambling company, funded an entire twitch competitor - Kick - just to bring back gambling streams. They also pay popular streamers to gamble on kick. The streams themselves are apparently entertaining enough to draw tens of thousands of viewers per stream, which is comparable to e.g. many millions of youtube views. To me, it's strange - you're just sitting there as slots, tiles stream by and sometimes match up, and someone else's finances do a random walk. (While the losses the streamer makes are real in one sense, it's more than compensated for by their pay). And the streamer isn't even pulling a 'lever' with their finger, it's fully automated. (As this vod demonstrates - letters and colors fall for a full hour next to an empty chair). I understand the appeal of watching the talented play sports and esports. I understand the thrill of gambling. But when you've taken out both the skill and the risk, what's even left?
curious_straight_ca
8d ago
Books are literally free if you have a computer or a phone. I mostly use annas-archive.org now, it has libgen and zlib's contents
curious_straight_ca
8d ago
I strongly disagree, and think that this disagreement is a 'crux' for the race-iq disagreement. Obviously if genes play no role in individual intelligence for the median individual, it plays no role in group difference.
Otherwise the "noise" of individual environmental factors is inevitably going to overwhelm whatever "signal" there is that can be drawn from genetics
You are/were engineer of some sort, right? Imagine you're buying parts for a series of machines at a factory. One brand fails in an average of 3 years, another brand fails in an average of 4 years. These are the only two possible choices, and they both have the same price. But - the standard deviation is .7 years! Does the "noise" overwhelm the "environmental factors"? In one sense, yes, for any individual washer it's tough to guess which brand they came from. In another sense, no, you know which brand you're choosing. And if you see a washer last 7 years, its' probably from the second brand (depending on the distribution).
The point is that whether the environmental 'noise' washes out the gene 'signal' depends on how large they both are. If genes cause +.1 stddev, then they don't matter in practice. If they cause +1 stddev, they do. Different natural histories for humanity could give us alternatives either way. If we took my gene editing suggestion - in a hundred years - probably environment would suddenly become the best way to improve human intelligence again (well, ignoring AI). But we have to measure how strong both are.
And that takes us to science.
Intelligence is a core construct in differential psychology and behavioural genetics, and should be so in cognitive neuroscience. It is one of the best predictors of important life outcomes such as education, occupation, mental and physical health and illness, and mortality. Intelligence is one of the most heritable behavioural traits. Here, we highlight five genetic findings that are special to intelligence differences and that have important implications for its genetic architecture and for gene-hunting expeditions. (i) The heritability of intelligence increases from about 20% in infancy to perhaps 80% in later adulthood. (ii) Intelligence captures genetic effects on diverse cognitive and learning abilities, which correlate phenotypically about 0.30 on average but correlate genetically about 0.60 or higher. (iii) Assortative mating is greater for intelligence (spouse correlations
0.40) than for other behavioural traits such as personality and psychopathology (0.10) or physical traits such as height and weight (~0.20). Assortative mating pumps additive genetic variance into the population every generation, contributing to the high narrow heritability (additive genetic variance) of intelligence. (iv) Unlike psychiatric disorders, intelligence is normally distributed with a positive end of exceptional performance that is a model for ‘positive genetics’. (v) Intelligence is associated with education and social class and broadens the causal perspectives on how these three inter-correlated variables contribute to social mobility, and health, illness and mortality differences. These five findings arose primarily from twin studies.
Essentially, when you try to measure whether the signal or noise is higher with twin studies, the signal is stronger. We can also do GWASes, where we estimate the effect of individual genes on a trait like intelligence. Each gene has an extremely small effect, but when put together this can predict (some of the time - genetics-based heritability estimates are lower than those from twin studies, because many genes + small effect is hard, but it's increased over the past five years) which of two siblings will score higher on intelligence tests just from their genes.
This is also intuitively true. Clarence's son with another lawyer will be smarter than random joe. If Mahomes marries the daughter of another great football player, they'll be a better athlete than joe.
Chesterton, What's Wrong With The World
It's a long series of short and funny meditations on politics and philosophy. A typical paragraph:
I disagree with many of them - e.g. the above hints at distributism, the idea that individuals will make better use of their property if they own and work it independently, rather than a few capitalists owning it all. But the capitalists mostly make more efficient and productive use of it, which is why individuals sell their productive capital to capitalists - the capitalists can pay more for it than the individual would make on his own. But it's still funny.
More options
Context Copy link