Put another way: the economy needs Dalits; there must always be sewage janitors working on minimum wage.
I would argue that this has become less true throughout the industrial revolution.
In ancient times, almost all of the population had to engage in backbreaking agricultural labor, with most of the surplus going to military aristocrats. Today, even minimum wage jobs typically require literacy and are a lot less back-breaking than working the fields.
You can argue that instead of exploiting our own poor, our wealth is built on exploiting the poor of other countries, but I do not think this is substantially correct either. Of course having countries where labor is cheap which produce electronics or textiles increases the wealth of the Western world, but fundamentally there is more to the success of Western countries than just exploiting the Global South.
The job market for shit jobs has some supply elasticity. If unskilled labor costs next to nothing, automation is often not worth it. On the other hand, if unskilled labor is scarce, then using automation is a great way forward. A wagon driver can probably replace ten porters, and a truck driver can replace plenty of wagon drivers.
In medieval cities, you needed to employ a lot of people to shovel all the shit. You need vastly fewer people to run a modern sewer system. OTOH, the job of sanitation janitor is much more complex than simply shoveling shit. Depending on the supply of labor, I would not be surprised if they made actually more than minimum wage in richer countries.
I think that adding anti-tampering devices to nukes is feasible. If tampering is detected (for example if the internal pressure changes), a nuke could disable itself by selectively firing a few of the explosives, spreading the fissile material over a few meters. Scraping off that material and building a nuke from it is doable but much harder than just replacing some software. (Of course, I would go for a few-kiloton fizzle in the event of tampering.)
How many of them really, truly want to kill tens of thousands at once?
Your classic bomb-throwing anarchist, commie terrorist or car-bombing separatist might shy away from nukes. However, religious extremists are different. I think the reason that 9/11 did not explode a nuke in NYC was not that they wanted to minimize casualties, but that they did not have a viable path to getting a nuke (or getting it into NY).
Perhaps Hamas would not use a nuke against Israel (not that I would bet on it), but the fact that a significant portion of the Gazans support them indicates that there is likely a more radical fringe.
The point of strategic nukes is not to get advantage in the situations where you use them. The point is to decrease the probability that it comes to that point, to ensure that the need for their use remains counterfactual.
A general policy of "when faced with an existential threat, such as an invasion, we will nuke cities of the aggressor" will do fuck-all to stop an invasion. However, if you credibly pre-commit to following through on it, the chances that you get invaded in the first place will be much smaller because most countries do not consider the glassing of their cities an acceptable price for waging war.
Also, you can not invade to stop a country from using nukes. The time scales for launching nukes are in the minutes, the time scales for invasion are on the order of days.
I like the proposal to have the universities act as guarantors of the loans of their students. They have control over how employable their students will be, over how much debt they will accrue in tuition and over how long the ones who will fail will stay on before being failed.
If a university then decides that that they can not provide a course at a cost where the students can on average pay back the costs, then it seems reasonable to conclude that their course is not a sound economic investment. They might still attract students who are independently wealthy and just want to study something for fun, but there should be no student loans for such courses.
Of course, over here in Germany, we don't have tuition. Bafoeg, our student loan system (used by 20% of the students) is capped at some 900 euros a month (most of which will go towards rent, in the big cities), and you have to pay half of it back. Universities don't have famous sports teams here.
I think "disparate impact" is a ridiculous standard. The odds that any decision process will return the same results for two groups which differ on all sorts of socio-economic axes seems unlikely.
I mean, I could get behind that some uses of ML in some fields might be unfair. For example, there might be rational economic reasons to discriminate against certain minorities. If Mormons are 10000x more likely to be killed by bears (because bears are murderist racists or something), then it might make economic sense to not hand out loans to Mormons in bear country. Even if the religion of the applicant is not explicitly present in the input data, a neural network could just learn to infer "applicant is a Mormon" from all sorts of proxies like name, place of birth and so on. If we disallow bank directors saying "no Mormons", we should also disallow such NN for consistency. By contrast, just indirectly discriminating against Mormons because their financial situation is worse (perhaps due to all these bear-related funeral expenses) would seem fine to me even though it has a disparate impact.
While it may be useful to force the market away from the economic optimum in certain situations, the idea to apply this to everything seems profoundly silly. If I (male, 40, overweight) were to post nudes on OnlyFans (not that I intend to do so), I am sure that between user rankings and their recommendation engine, I would end up making a lot less than the median OF model. That is a disparate outcome from an algorithm right there. Should I be able to force OF to push my pictures more?
Or say someone decides to run their blog in French because they have "limited proficiency in the English language". Should Google search be allowed to filter that result if people search for English language websites? That is a disparate impact right there!
I for one would like to apply stricter standards to the prosecution.
For example, the fact that the defendant can misrepresent the facts as much as they want is not a good reason to also allow the prosecution the same leeway with the truth.
The standard for criminal trials is generally "beyond reasonable doubt". If A and B commit a crime, and prosecutor X convinces their jury that A was the mastermind, and prosecutor Y convinces their jury that B was the mastermind, and they both worked from the same evidence, then at least one of the prosecutors is grossly miscalibrated about what reasonable doubt should mean.
Prosecutors who try to convict people of of stuff they are actually guilty of will not show this behavior. OTOH, with this lower standard, you could have multiple persons be convicted for having fired the same gunshot.
If there was a credible prophecy that US companies would develop ASI in twelve months, then invading Taiwan (which would result in the TSMC fabs getting destroyed, and the capability of Nvidia to build new AI chips being lowered) might be a hail mary that the PRC could try.
Absent such a prophecy, risking a substantial chance of WW3 to slightly delay a potential singularity does not seem the best survival strategy.
First we have to ask whether China even wants Taiwan.
The interesting thing about Taiwan are the state-of-the-art TSMC chip fabs, which are better than what the PRC has.
However, these are fragile things easily destroyed in the event of an invasion, and keeping them running without support from ASML would likely be hard to impossible.
And while Taiwan is important, it is not the only place in which the US could manufacture state of the art chips for military use, so taking Taiwan would hurt the US economically but not cripple it militarily.
Of course, Taiwan might also be a Schelling spot for anti-Communist Chinese, like Hong-Kong was before. But risking World War 3 to drive your ideological opponents from Taipei to San Francisco seems rather pointless -- especially if you can also just impose Honecker-style controls on your people so they can not defect.
Indirect control. China implements air and sea border controls to make Taiwan a self-governing administrative region of China. There is no need for a direct attack on Taiwan or any blockade of usual commerce. Without initiating violent action, the Chinese can assert sovereign control over the air and sea borders to Taiwan, establishing customs and immigration controls. This is not the same thing as a blockade. A blockade would instead become one of the possible consequences if the other side violently challenged China’s assertion of indirect control.
Establishing "customs and immigration controls" on any territory outside your jurisdiction, such as the high seas is exactly what a blockade is. I think that the UN SC would be unlikely to okay a blockade of Taiwan, which would make it illegal. In the end, a blockade is dependent on your willingness to shoot at blockade runners. Shooting at blockade runners in the case of a blockade not sanctioned by the SC is an act of war.
Also, an important difference between go and the real world is that in the real world surrounding someone is not sufficient for capturing them. Cutting off Taiwan from the rest of the world will not cause them to raise the PRC flag. I think it is likely that the food situation (production-consumption-ratio) of Taiwan is better than that of Gaza. Besides, as I learned when discussing Gaza, International Law kinda says that you may not declare food contraband.
First, EMPs, super or otherwise, are thermonuclear weapons. using them in an opening move poses an even more serious risk of things escalating quickly to global thermonuclear war than they would with a conventional attack. To my knowledge, there is no principle in international law where EMPs do not count as a nuclear attack.
Second, Faraday cages exist. I am not an expert, but I would expect that the electronics within a tank can be well shielded. The thing which might be hit is stuff which intrinsically has a link to the outside world, such as radio antennas and sensors (especially em sensors such as radar, but possibly also cameras).
Third, even if you manage to fry all the enemy antennas and radars before an invasion, the obvious countermove would be for the US to also cause an EMP mid-invasion. If the US don't have super-EMPs, conventional hydrogen bombs could do the trick -- you can always compensate by being closer to your target. This would largely level the playing field again.
Well, it looks to me like Russia has some 200km of undisputed waterfront on the Black Sea. Wikipedia lists two ice-free major ice-free ports: Novorossiysk and Taman. I am sure that for a fraction of the price of that special military operation, Putin could have gotten a top-grade port on his coastline.
Also, I am not really condemning Russia too much for sizing Crimea. That operation seems to have been accomplished with minimal bloodshed, at least. My main problem with Putin is his behavior since 2022, when he tried to take Kiev (ca. 250km from the Black Sea, not a great location for a port) in a surprise attack, and opted to fight a war of attrition when this initial attack failed.
Also, I think there can be some debate on in what cases the security interests (legitimate or otherwise) of big polities trump the right of sovereignty of smaller polities. On the one hand, if Lichtenstein entered a military pact with North Korea, I think that the rest of Europe would be correct in denying Lichtenstein the opportunity to station North Korean nuclear missiles in the middle of Europe. On the other hand, I don't think most US meddling in South America was in pursuit of legitimate security interests. Meddling in Panama, Mexico or Canada, as well as the Cuban missile crisis are somewhere in between.
Now, I am not a US interventionism fanboy. I believe that a lot of the military and CIA ops the US engaged in the cold war and the Bush II era were net negative from a thriving of humanity point of view.
But writing from what cynics would call a US client state but what I prefer to see as a minor member of the status-quo coalition (Germany), the USA makes a pretty decent hegemon (in Europe, at least). The values which they prefer (market economies, free trade, individual rights) seem to work out better than what other local hegemons have enforced in part of Germany before.
There is a reason why a lot of countries in the former Warsaw pact wanted to join NATO instead of forming a defense pact with Russia against NATO aggression -- they had just spent a few decades at the receiving end of such a defensive pact.
If you take the right to self-determination of peoples seriously, then there can not be a right to preemptively conquer weaker neighbors to prevent them from joining defensive pacts against you.
Also, who in their right mind would want to invade Russia? Europe tried it twice, with disastrous results. Invading a major nuclear power is not a decision anyone borderline sane would ever consider in earnest. What Russia is defending by invading Ukraine are not legitimate security interests, but their status as a local hegemon who can use force against weaker neighbors at their discretion.
The US is not to blame for all the evil in the world. There have been wars for millennia before the US was a thing.
Regarding that Sylt video, what struck me most was how badly performed that act of rebellion was on a factual level. What is wrong with generation TikTok if they can't even LARP at being Nazis better than that? "Auslaender raus"? The 90s called, they want their NPD back. Today it is "Remigration von Bundesbuergern", which is way more offensive btw. Doing the Hitler Salute while also indicating the Hitler mustache with the other hand is something I have never seen outside of people making fun of Nazis, and would likely have gotten you into trouble in the third Reich.
Anyone who has two brain cells to rub together, went to school in Germany, and decides to annoy the mainstream by voicing Fascist thoughts should be able to do better than that. I mean, I don't expect a torch march while they sing the Host-Wessel-Lied, but would it be too much to ask that they perform the salutes correctly? I sure hope Hitler has youtube access in hell so that he knows what became of his ideology (not that it had not started out with drunk losers in the first place).
I am not saying that I approve of that kind of rebellion. It is like seeing a four-year-old who decides that he will annoy his parents by smearing dog turds in his face and then falls down as he tries to pick the poo up.
And just like I would not want that four-year-old being on national TV, I also don't think the Sylt people should be. From the way they act it seems likely that they are not what one would call satifaktionsfaehig (capable of giving satisfaction, i.e. in a duel). Now, if one of them was a leader in a youth group of a political party, that would be a reason for this video to make the news. And if this is part of some weird TikTok trend where this is among the top ten things in Germany, it would be fair to report about that.
I mean, if a court decides that they violated the law and makes them pay a fine, I am totally okay with that. But I don't like like mobs, and there will always be some who will see this as their opportunity to prove what valiant anti-fascists they are.
This would also provide bad incentives, because it would cap the risk of decisions with huge negative externalities. 10% of revenue times the probability of getting caught is basically nothing, so unless your action is going to cause a big enough stink to move Congress to act, you are in the clear.
As an analogy, suppose we capped the fines and damages for gross negligence of humans at 10% of their annual income. This would provide terrible incentives: people could speed by near kindergardens, throw empty glass bottles from skyscrapers, operate on patients while drunk and the like secure in the knowledge that the worst outcome will cost them no more than they spend on vacations.
Corporations already have huge advantages over natural humans through diffusion of responsibility and their liabilities being mostly limited to their assets (so the risk to their investors is limited). For Thalidomide, the corporate death penalty (i.e. bankrupting a company through fines and damages) seems like an appropriate outcome.
Of course, glyphosate is very far from Thalidomide, but caps on damages are not the answer.
I think simply embedding a copyrighted artwork in junk DNA will probably not fly in court. But if software can be copyrighted, I don't think there is a reason why GMOs -- whose design also involves some stylistic decisions -- can't be. I hope I am wrong about this, though.
The reason that patent law is widely considered more evil than copyright law in software is that avoiding copyright infringement is often a very easy task (the exception being APIs and the like). If you have a copyrighted bubblesort library, I can simply decide not to use your library and write my own implementation. By contrast, if you have patented bubblesort (depending on the sanity of the patent system in question, you might have to patent a specific instance of bubblesort which affects something specific in the real world), then every other implementation is infringing.
punk bands
For some reason the idea of punks of all people forming rigid, hierarchical societies which have palace intrigues and make sure that their members conform to their norms feels really amusing in an absurdist way to me.
At the moment, the US attracts people who see their country of origin going nowhere and are willing to emigrate to secure a better future for themselves and their kids. Due to an oversupply of such people, the US can -- in principle -- filter for the best and the brightest of them.
What you are offering is instead is an oil rig job -- hard work, hostile environment, with the only motivation being able to spend the shittons of money you made after you come back to civilization.
Now I am sure that you would find takers for that deal, global income disparities being what they are, plenty of people would jump at the chances to pick up dog turds for minimum wages which will allow their extended families to live comfortable lives. I mean, it is not like the Arab oil dynasties have problems finding wage slaves either. Put in the hours, go back home, live a better life.
But the top of the cream will likely look elsewhere. Why would anyone take a professorship in a country which has made it plain that they would kick them out as soon as they retire when they could go to Canada instead? Likewise, there is certainly the trope of immigrants who are working hard so they can fulfill their dream of some day owning a Kwik-E-Mart.
The lower class of US citizens will not much like the outcome of your proposal either. Suddenly they have to compete against people who are completely beholden to their employers for continued residency and may come from cultures in which unions are not a thing, while also willing to work for much lower wages because they do not have to feed a family in the US from them. If you think Facebook's preference for H-1Bs over citizens was bad, wait until Amazon gets to staff all their warehouses with people who have much worse visa conditions than H-1B.
I also have some moral objections to your proposal. Leaving aside the question if indentured servitude a la UAE is really the path the US should follow, I also believe that people should generally be the citizen of the country they have lived in for generations. Your proposal would create a permanent caste of people who are non-citizens. Given the TFR, it seems entirely possible that at some point a significant fraction of the population will be excluded from democratic participation. At some point you in effect have an aristocracy. This feels deeply un-american to me.
And if you go back far enough, every human outside Africa is an immigrant from Africa.
I think one strategy is to pursue mainstream policy while also convincing your base that your opponent would totally end democracy.
You are technically correct. However, Monsanto was acquired by the pharma giant Bayer, who decided to discontinue the Monsanto brand. If instead they had gone bankrupt or be acquired by a company which imposed a drastically different business model, things would be different, but this looks to me like an acquisition followed by a corporate rebranding while keeping the same business practices.
In a similar vein, I will continue to say "acquired by Google/Facebook" instead of "acquired by Alphabet/Meta", "posted on twitter" instead of "posted on X", "addicted to heroin" (which is a trademark which has not been used for almost a century) instead of "addicted to diacetylmorphine", "Blackwater" instead of "academi" and so on.
It is linked under library on lesswrong, so it is kind of the top of the canon as far as rationalist fiction goes. Other major works are:
- Scott's Unsong (Original setting)
- Alexander Wales' "Worth the Candle" (Original setting)
- Alicorn's "Luminosity" (Twilight 'fan'-fiction, of all things)
- Eliezer et al's planecrash (collaborative glowfic, Dath Ilianian isekaied to the world of the RPG Pathfinder)
All of these authors also have shorter fiction. If you are unsure, I would start with something shorter and see if you like the style. There are also audiobooks of HPMOR and Unsong.
I think "someone says something with is laughably wrong and also outrageous on the motte" does not warrant a top-level post.
We have our share of trolls and Nazis. Every once in a while we get an holocaust denier. I am sure that I could make a civil argument about why we should kill and eat all the people with the letter x in their legal names as a matter of national policy without getting banned for it.
I think it is important to establish that a lack of refutation is not the same as silent approval. Trolls can make outrageous statements a lot quicker than anyone can refute them. See xkcd.
I mean, mods could use the "Proactively provide evidence in proportion to how partisan and inflammatory your claim might be." rule to shut down such comments, by ruling "denying humanity to a significant portion of extant h. sapiens is a claim which is so inflammatory that you should require ten recent articles in top-notch scientific journals making that claim as evidence" (and then relying on the fact that this level of evidence can not be met).
But this would also put us on a slippery slope to tone policing. As the late Niemoeller observed (paraphrased): "First they came for the holocaust deniers, and I said nothing because fuck Nazis. Then they came for the blacks-arent-humans people and I said nothing, because fuck racists. Then they came for the genetic-IQ-difference crowd, and I said nothing because it was not a topic which interested me. Then they came for me for using a generic masculine form and nobody was left to speak out for me." So I am kind of fine with the odd Nazi comment as a price to pay for having a forum where any opinion can be expressed, because some other controversial opinions are at least interesting.
How about a culture war outside the traditional red/blue conflicts for a change?
The Guardian ("I read it for the math problems") reports on the decision by a court in the Philippines to ban golden rice, a GMO plant designed to combat vitamin A deficiency. The NGO arguing for a ban (aside local farmers) was Greenpeace.
[https://www.supportprecisionagriculture.org/nobel-laureate-gmo-letter_rjr.html](168 Nobel laureates) have called on Greenpeace to stop campaigning against golden rice in 2016. Here is a discussion on EA forums.
Now historically, I have not been vehemently opposed to Greenpeace. When I was a kid in the 1990s, they were protesting French above-ground nuclear weapon testing, which seems fair enough. While pro-NPP myself, I think there are solid arguments to be made for opposing nuclear power (like proliferation risk and long-term disposal of used-up fuel elements), but I can't understand why you would target NPP before you would target fossil fuel power generation.
The steelman of the Greenpeace argument would be that allowing patent-encumbered GMOs will be a foot in the door for pushing more GMOs on rural farmers which will eventually result with Monsanto owning the small farmers. The situation for GMOs is not unlike the situation for software: expensive to develop, but cheap to copy. As a free-software advocate, I very much would prefer outcomes where the companies who develop the software/GMOs do not end up with a stranglehold on the end users due to copyright or (even worse) patent laws.
At least for software there exist alternatives like FLOSS. From reading the FAQ of golden rice, it looks like they could not develop their product without using technologies patented by biotech companies, so they got to them to agree to waive licences for farmers who make less than 10k$ per year, which is their target audience. This is still far from ideal (better would be a blanket free licence for golden rice, or constraining biopatents so much that you do not have to ask Japanese Tobacco to licence your rice plant, or perhaps abolishing them altogether), but does not seem like a terrible deal -- especially if you have a local court system which is rather pro small farmers.
So my conclusion is that Greenpeace's opposition is unreasonable and they have been become one of these organizations who advocates for policies which are deeply unpopular (like PETA, or "always believe the woman" groups) in the wider population as members race to signal how committed they are to their cause.
There are states where ethnicity matters and linearizing different interest groups will (a) not always be possible and (b) result in some ordering which is drastically different from from the general usage of "left" or "right".
I mean, look at the Knesset. I am not an expert in Israeli politics, but Wikipedia describes Ra'am as "an Islamist and conservative political party". They sit on the far left, but apart from ethnic concerns would probably belong right of the center -- where none of the Zionist parties would have them in a coalition.
(Of course, another anomaly would be the dirty trick when labor tried to form a coalition with the ultra-orthodox, but you can argue that the fate of that attempt is mostly proof that cutting out a middle party on the political spectrum does not work.)
In Germany (where I know the politics a bit better and they are simpler -- the Bundestag does not have a zillion different shades of blue and red like the Knesset on Wikipedia), a counterexample would be the Grosse Koalition.
German major parties go left to right (WP Bundestag colors in parenthesis):
(This is how they are arranged in the Bundestag, there is a case to be made that the Greens are actually left of the SPD.)
Take the 2005 Bundestag. Nobody wants to form federal coalitions with the Linke or the AFP (so far), so the following would get you a majority:
Per your "arranged by compatibility" argument, one of the first two options would be favored, covering as little as the political spectrum as possible. Even if you agree with me that the Greens really ought to be placed on the left of the SPD, an "Ampel" coalition would seem preferable.
The reason that they ended up having a grosse Koalition instead was that while seated next to each other, FDP and Greens had major policy differences. These differences are not well reflected on the traditional left-right-spectrum. The Greens wanted to shut down nuclear power plants, impose a speed limit on the Autobahn and generally have stricter environmental standards. The FDP wanted none of these. In the end, it was easier for SDP and CDU/CSU to compromise than to reconcile FDP and Greens. Three out of the four Merkel cabinets were such coalitions (the other was CDU/CSU+FDP).
To be fair, getting the FDP into the coalition boat would have been possible policy-wise without to many concessions, but other considerations made this unfavorable. (At an election, the previous administration is mostly not seen positive. Having the stink of culpability in the failures of the previous administration on you has to be balanced with being able to point your constituents to specific policy wins your ministers accomplished. This means that you generally go for the minimum viable majorities -- and the grosse Koalition already had a solid majority.)
More options
Context Copy link