If you've noticed that Boeing has been in the news quite a bit recently, and not for the reasons they would like to be, you may be interested in this article.
It has some interesting things to say about the "Ricardian fallacy," and treating financial reports as the be-all-end-all of understanding the state of a business.
I remember reading recently... not sure if it was on Less Wrong or Astral Codex Ten... a great insight about colonialism: how it made the colonies poor, *without* necessarily making the empires rich. The colonies suffered a lot and had their wealth stolen, but the empires had to build armies which cost a lot of money and it is not obvious that the entire thing was a net benefit. So instead of modeling the situation as "X stealing money from Y", it would perhaps be better to see it as "X burning Y's money".
And I just realized how all this logic also perfectly applies to Soviet Union. Soviet Union made its satellite states poor *without* making itself rich. An average person in Russia was actually much poorer than e.g. an average person in Czechoslovakia. So from the perspective of the person in Czechoslovakia, it made sense to complain that their country was made poor by the decades of Soviet occupation... and from the perspective of the person in Russia, it made sense to object "what do you mean by 'poor'? look at me -- if you think that I have your money, you are crazy".
The answer is that there is no contradiction between "X making Y poor(er)" and "X not becoming richer in the process". Just like in real life, breaking someone else's legs doesn't make your legs healthier. But with wealth, people often have this fixed-sum intuition, and expect that if wealth was taken from somewhere, it *must* have appeared somewhere else (and if it did *not* appear somewhere else, then the person complaining about the theft is lying).
That makes me think about the later development of Eastern Europe. How much of the economical growth was the financial support from EU, and how much was... simply the Soviet Union not ruining things anymore. (This is not a dilemma; I assume that *both* numbers were huge.)
I’m not an expert at UK law, but I read that as two years in jail and an additional 2 years suspended, not the same 2 years suspended. If you and OP are right, that’s a valid point, but the more natural reading is that those are consecutive.
My project recently released this - it has been described as a return to OK Go style, true effort music videos - Inception meets Fear and Loathing, it imagines how democratised neuralink-like tech could unfold.
In the UK, you will *literally* receive a harsher penalty for distributing "racist" stickers than for RAPING a child (well, at least if you're a brown rapist, not sure about if you're white).
I'm trying to find a specific SSC post from years ago. In it, I remember that Scott defines what "common knowledge" is, and something about how it involves an infinite regression of that everyone knows that everyone else knows that everyone else knows something is true. I don't think it was It Was You That Made My Eyes Blue, I think it came earlier than that. I remember maybe something about a teacher asking kids to stand up in a classroom...? Any help appreciated. Thanks!
As someone who initially thought that Rittenhouse was probably guilty but eventually came around to it being legitimate self defence, I do think that I keep an open mind. And thus it's important to highlight when I have a major update in the *other* direction as well. I'd never paid much attention to the Floyd case before and had gotten an inaccurate picture due to hanging out in conservative circles.
Last night, I read this post about this case, and the details are far, far worse than I'd imagined. It really is appalling. And thus I'm reposting this in the hopes that everyone else reads it as well.
Does anyone else worry about the concept of online "bait", and how it's becoming more common to ignore anyone with a wildly different view or opinion?
Any idea that's sufficiently idiotic or ill-informed seems to be immediately written off as 'someone trying to mess with people for a reaction', and deemed not worthy of your time. But it seems perfectly plausible that there could be someone unironically touting such views, and in that case, do we not just reinforce that person's echo chamber by choosing not to debate them? Where do you draw the line?
I ask because I see plenty of obvious bait on a daily basis - that's not the stuff I'm referring to. Rather, I mentioned a post in this very comment section, and my partner immediately said "ha, and people were actually replying to them?" like it was an impressive feat of trickery, but the vibe I got was way more sincere. If we start treating every "crazy" opinion we see as "bait", then are we not just siloing the people who honestly hold these views?
Forget China, Japan, and Europe. I just discovered that *Uzbekistan* of all places has a bullet train. And they built it in 2011, in only seven months, less time than it takes an American builder just to scratch their ass. Now I'm really feeling civilizational inadequacy.
For those who like mixing good deeds an capitalism: there is a hole in our financial system. We have banks to fund mundane businesses. We have Wall St. to fund huge corporations. We have Venture Capital to fund businesses which have the potential to become huge in a reasonably short period of time, or be sellable to a huge venture.
But what of creative businesses which have moderate, but nontrivial risk, but don't have the potential to become Wall St. sized? We need an equity market for such businesses so VCs and angels can cash out. Here is a proposal for such a market:
Does anyone have any recommendations for "political" novels? I don't mean defending any particular political system or policy. I more mean a novel of factions and scheming and plotting and stuff. It doesn't even have to be in the arena of actual government. I'm also super genre-agnostic - though I've been on a sci-fi kick lately.
Basically, I suggested that something like the Radbruch formula be applied to historical laws backwards pretty much indefinitely, rectifying all unjust transfer of property as much as possible.
Specifically, I suggested the following rough test:
"a) Was the transfer of property in question legal at the time?
b) If so, were the institutions of the time that governed the transfer morally or legally uncontested at the time for reasons widely accepted today?
c) If so, can these institutions reasonably be considered just when viewed in light of fundamental rights that ought to hold at any time or place?
d) If the transfer in question fails tests a,b or c, can the plaintiff prove a causal connection from the unjust transfer of property to present outcomes?
If so there should be reparations, regardless of how much time has passed since the event."
This being a rather radical view, my proposal drew a fair amount of criticism and ridicule. This came mostly in the form of putative counterexamples which my approach would not be able to handle or in which it would yield absurd consequences.
Suppose we found out that Ea-Nasir indeed never refunded Nanni his money and some descendant of Nanni would bring a case against some descendant of Ea-Nasir today. Suppose further that Ea-Nasir indeed defrauded Nanni according to the laws of the time failing part a) of the test. Finally, assume that the standard of proof to be met is "preponderance of the evidence" which is usually required in civil cases.
Reply to 1) Plaintiff's lawsuit would fail because they would almost certainly fall short of meeting part d) of the test, which requires proving causality.
This is for several reasons:
i. Dilution: if Ea-Nasir has any living descendants at all, then approximately everyone on earth is his descendant (https://www.theguardian.com/science/commentisfree/2015/may/24/business-genetic-ancestry-charlemagne-adam-rutherford). The same goes for Nanni. So approximately everyone could appear on both sides of the lawsuit and the amounts owed would be infinetisemal. Any possible causal influence has been diluted and spread around among both parties descendants so much that it would almost certainly be a wash.
ii. Non-ergodicity of economic growth: while global GDP has increased manyfold since ancient Babylonian times, this does not mean that individual fortunes have. Time average and ensemble average of the wealth building stochastical process diverge. In particular, fortunes decay over time (and symmetrically do debts): conventional wisdom says that fortunes are built in one generation and lost by the third generation. While this belief has been challenged (https://cdn.ymaws.com/my.ffi.org/resource/resmgr/docs/goodman_study.pdf; although I will in a later post present some evidence that the three generations rule may not be far off after all), it remains true that fortunes that are more than a few hundred years old are exceedingly rare. This is surely not unrelated to the fact that debtors' heirs can usually simply refuse to accept the inheritance.
iii. Lack of data: obviously.
In light of this, any suit being brought that is based on claims older than a few hundred years is almost certainly going to fail. I would expect that after a change of the law a few such lawsuits would be attempted largely for publicity but then rather quickly fizzle out.
Reply to 2) The value of ancient claims would not skyrocket to astronomical heights. Insofar as one can establish causality, one would also likely get an estimate of the effect size. For the reasons already stated, it is not likely to have increased dramatically through the generations but instead would probably have decreased. For this reason a fixed rate of interest is simply completely unsuitable to simulate the development of the value of a claim over long time periods. The interest rate is meant to reflect the counterfactual development of the value of the property of which the plaintiff was illegally or unjustly deprived. But empirically speaking people don't put their money into a savings-account, ETF or similar over many generations. Rather they consume it, make bad investment decisions or just get unlucky. Moreover, until a few hundred years ago GDP-per-capita was essentially flat (https://ourworldindata.org/grapher/global-average-gdp-per-capita-over-the-long-run). In such time one could not expect one's assets to appreciate over time even if carefully managed. For these reasons, courts would very likely not use a fixed interest rate to determine the present value of a claim but would rather estimate it from the causal evidence that the plaintiff has provided.
This was only the first worked example, but it is getting late here. Tomorrow I will try to post one on slavery in the US. That one will in particular include some concrete values.
Anytime I get through reading a Links post (not here necessarily, just any listing pointing to lots of Substacks), my primary thought is "man, it annoys me how many blogs won't let you Like a post without first subscribing". Doesn't matter if it's a free tier sub, Beware Trivial Inconveniences. Turning off comments to non-subs makes sense for the same reason, but...what, it's important to prevent frivolous drive-by likes from the uncommitted? I guess? Seems like a bad deal for new blags hurting for exposure, which tends to be what ends up on such roundups in the first place.
Semi-relatedly, I swear that sometimes Substack loses track of historical likes. Will occasionally reread something I'm certain I liked (both senses) a year ago or whatever, and that heart is just...gone. Which is frustrating, since an existing heart serves as a good bookmark that yes, I did in fact read this before, and can check it off the list after a brief skim.
I'm curious if anyone in the community works in the GIS/Geospatial field. I'm about to write my bachelor's thesis in Urban Planning, and have become interested in a career in GIS. I've browsed https://80000hours.org/ for GIS work that might have a significant positive impact, but come up short.
I'd be grateful if anyone would like to chat/give some advice on three main questions:
1. Whether a career in GIS/Geospatial is worth pursuing.
2. Potential research topics for my bachelor's thesis.
3. Whether to pursue a master's degree or start looking for work after I've graduated.
How is the accelerationist view coming? We went from ChatGTP 3.5 (2022-11) to ChatGTP 4 (2023-03) in four months, but now twelve more months have passed and I think it's fair to say that there has basically been 0 progress in text-based AI since then? No GTP5, lots of new competing models but none that are widely believed to be superior to GTP4. Sora was touted as a game-changer for video AI when it was announced, but it has no official release date and we have little info on its average quality or cost to run.
I realize this comment could age like milk any second, but is the current evidence consistent with a view that things are plateauing or at best advancing linearly rather than exponentially? One response could be that it's zooming in too far to compare the last few months to the previous few months, and indeed, the next few months could certainly show another spurt of progress. But on the other hand, isn't the concept of accelerationism / exponential growth that the curve is the same no matter how you zoom in, and progress durations are supposed to get shorter and shorter, not longer and/or the same?
The article mainly argues that the USA is "better at being Israel than Israel", in the words of the writer. That is, the USA's population of Jews is 85% of Israel's, and they have a better quality of life across a broad spectrum of metrics than their Israeli counterparts. The USA also absorbs Jewish immigrants with less hostility and less friction than other groups, including - first and foremost - Arabs and/or Palestinians and/or Muslims.
Article aside, this section got my attention:
>> Defenders of the modern Israeli state are in this constant argumentative bind: they must ceaselessly insist that Israel is teetering on the brink of destruction, in order to keep American money and weapons and diplomatic muscle flowing, while at the same time claiming that Israel is the only place where Jews can be safe. These are, obviously, directly contradictory sentiments. If it takes the constant patronage of the most powerful nation on earth to keep Israel from destruction, and even then the country is subject to assaults like that of October 7th, in what sense could Israel possibly be considered a safe place for Jews? Well, you could point out that by most basic statistics that catalog violent threats, Israelis are safer than citizens of most countries in the world. (If you ask an Israeli whether their country is safer than Chad or Colombia or Pakistan, they’ll get offended that you asked.) Unfortunately, you are then merely pulled back into the other side of the paradox - if it’s true that Israelis are remarkably physically safe, in context with much of the rest of the world, how can we justify the seemingly perpetual outlay of vast amounts of American ordnance and treasure on Israel’s behalf?
I think this is a general pattern in politics and similar ingroup-outgroup dynamics, not specific to Pro-Israel supporters, people simultaneously want to believe "The Enemy is Strong" and "We Are Strong". They selectively switch between those 2 according to which is most convenient at the moment, I have caught myself doing it several times in other issues not related to Israel-Palestine.
Anyway, the article shines a light on the many contradictions at the heart of Zionism: Israel is the *holy* land, Israel is a *secular* democracy. Israel is the *safe haven* for Jews everywhere, Israel is in *existential danger* and you can't hold a state to high moral standards when its existence is at the line. Most moral army in the universe, but don't hold it to higher standards than any other army, that's antisemitism. Israel is the only place where Jews can truly be safe, Jews in Israel are terribly unsafe (and the Palestinians are at fault).
Scott mentioned the problems with substack loading recently (much more of a problem on mobile for me). It seems like this is almost definitely due to the comments sections on posts all loading with the page, something that Scott's blog does differently from others, I believe because of Scott's requests. I know Scott doesn't want likes on comments so you can't just display the top comments, but is there some other solution, like only displaying the first 20 top level comments and then have a See More button after that?
I believe I remember Scott wanting to have the comments here replicate the way they worked on SlateStarCodex, but it seems like that's causing technical problems at this size of community.
On the morality of shooting Arsonists (And why we hate liberals):
A community near you sees two nights of it's businesses being burned down in the name of a political cause. In each case the extremely well funded government security force stands by indifferently while the very narratives which justified the attacks are propagated by government and press.
On the third day you show up armed to the perimeter of one of the few businesses that hasn't been destroyed. You like this business. It sold you your first car. It's a large business, a car lot so there's no possible way you can block all of it. All the cars in it are close together, so lighting up one means eventually lighting up all. A mob gathers and a man with a massive tank of gasoline and a zippo lighter screams, 'we are gonna burn this place down, nothing will happen to you.' You stand your ground and point your gun at him. He backs off, and runs. You believe he is retreating, but then you realize he's running towards another part of the massive parking lot. He pours the gas on a car, a prepares to throw the zippo lighter. You raise your gun, shoot him straight through the chest and kill him.
Was this morally justified?
---
The average natural right winger instinctively says yes. If you ask them why, most of them couldn't give you a great answer. But I'll give you a few now, each of which I've heard from relatively unintelligent people (if stated less eloquently):
- Property is the blood, sweat, tears and years of human life spent building it (ok normally not the blood part). To destroy it, is in a very real sense to kill a part of a person.
- You know better than anyone else what the value of your life is and if you think it's worth less than the pleasure you derive from destruction, I'm not gonna deny your assessment.
- I agree not to kill in exchange for the government agreeing to protect basic rights. If the government blatantly refuses to do so, only the laws of war and nature apply.
- A random business is not a legitimate target in a political conflict, destroying it has no value to whatever war the rioters are engaged in. They are war criminals.
(Of course, all of the above combined is it's own argument.)
I'm close to sure that there is at most a single digit percentage of educated liberals who take this position. I know this because I've gone over the public statements of any number of anti-woke liberals on the Rittenhouse case (a case infinitely more defensible than mine), and found no one who is willing to state unambiguously 'he was right to do what he did, including being there in the first place.'
Man in the liberal vision, is but property of his government or society; and if someone decides to destroy decades of a life's work and no one 'authorized by law' is coming to save him - well, it sucks for him. After all, it might, in some final utilitarian accounting be bad for society if a man took the law into his own hands - even in response to society's total abandonment of it. Of course, in any right wing vision, at least the non-religious ones a man who accepted this is no man at all.
Which is why, when push comes to shove; no compromise between liberals and right-wingers is possible. All talk of 'more in common' is empty. We believe that you would impose upon us norms that we could only wish on our worst enemies. That you might not personally hate us is of no consequence as we believe that you are fundamentally incapable of love for anyone at all. For all our enmity towards Communists and/or the woke, we can at least imagine a world from their perspective in which business owners are the real enemies who must be crushed to create a classless society. And boy, are they dedicated to this cause. Anyone has to admire that in the same way Orwell admired Hitler.
But for you, you who would condemn a man because he wasn't carrying a badge, because 'life is worth more than property', we have nothing but hatred.
Has anyone else noticed Google Doc's autocorrect get much more aggressive and stupider recently?
Previously, it would just attempt to correct mispellings or grammar errors. Sometimes it would get one wrong, but for the most part it was ok, and it would never attempt to apply the same correction more than once if you backspaced and changed it back. However, perhaps a few weeks ago, it got a lot more aggressive and also stupider, and now frequently makes changes that aren't even gramatically correct.
For example, when I was writing just now, in the space of one paragraph, it changed the "worn" to "worms" in "However, there were also small holes worn", and it changed "I applied" to "I apple".
Furthermore, it no longer gives up if you undo its correction. I've seen it apply the exact same (mis)correction three or more times when I backspace and fix the "correction".
Does anyone here know what's going on here? I'm curious what people on Google Docs would say about this.
I've been learning German for almost 2 years now. I've used a combination of in person classes, remote classes, and Duolingo. Across all of them, I probably average an hour-ish per day across those two years.
One thing I've noticed about Duolingo that I wish would be improved is that, while it occasionally will bring you old sentences with vocab that you haven't seen before, it's A) the same set of sentences you used when you initially used the word and B) still sentences where the entire context of the sentence is designed to help you figure out what the word is.
B is extremely helpful when you are first learning a word, but, in my opinion, it becomes detrimental when you are trying to cement a word that you already know, and prevents you from putting that word in a larger context.
I realize that generating a wide range of novel sentences for every vocab word, especially when those sentences would ideally have a decreasing gradient of contextual information, would be prohibitively expensive. This feels like the kind of problem that an LLM would be perfect for solving.
Language learning feels like the area of learning/education that is _most_ ripe for being completely upended by LLM assistants. I know that Duolingo is slowly trialing their "Max" AI assistant, but it doesn't seem like they are using it in an optimal way.
Duolingo is, for me, currently just a useful side thing that helps support and reinforce the learning I do in my classes. I think with a well thought out and intelligently implemented LLM assistant, it could become my primary learning tool. Unfortunately, I'm not sure that that would jive with their current incentives which lean towards gamification and just getting people to open the app every day. So Duolingo may not be the thing that does it, but man it feels like _someone_ is going to make the thing I'm envisioning, and soon.
For those who speak Chinese: I am looking for a name for a character in a story I'm writing. It does not need to be a word typically used as a name. It should be a term meaning something like "clever and brave." It should have a clearly positive connotation -- so it should not be commonly used to describe, say, talented criminals who are sneaky and sly. I asked on the Reddit Chinese language sub and got zero answers. Wow, Reddit's a lot less friendly these days! I'd be very grateful for suggestions here.
I'm trying to decide, Scott, whether you are the worst employment-opportunity signal booster in the world or the best....I will of course have a 1,000-word comment exploring this question ready soon.
(Which, in my ongoing effort to remain a sane person, I will then delete unposted.)
So as spring comes early yet again here in Western NY, I'm reminded of an observation*. Which is that global warming seems to be mostly a warming of the coldest temperatures and not an increase in the hottest temperatures. So in the northeast, winters are less cold and shorter, and the summers are longer, but don't really get any hotter... only in that there are more days with high temperatures.
There's been something pretty interesting going in philosophy blog Bentham's Bulldog: the utilitarian author has started to believe in God because it makes more sense than atheism. He doesn't hold theism with 100% certainty, but he sees it as more likely than atheism.
Does anyone here buy & sell individual stocks (or bonds, or crypto, or futures, or options, or commodities, or anything else?) I.e. does anyone here day trade, swing trade, or take longterm position trades? Anything other than just buying boring index funds and taking the market average. No judgement if you do, just totally curious- what motivates people to 'trade'? Are you profitable? More profitable than if you'd just held a boring index fund? Including taxes, which should take a significant chunk of any profits.
I spent a lot of time reading all the classic works on value investing, and I came to the conclusion that it's not worth the time investment for me to try to buy individual stocks and beat the market. But just interested in hearing peoples' experiences, good or bad
Anyone know exactly what the point of Austin's SXSW is? They have setious ppl speaking but also Meghan Markle speaking, they once had Barack Obama speaking, and I think, music concerts. You could listen or read these speakers' ideas elsewhere like on YouTube. I'm curious why anyone pays $2k to attend. I tried to find ways to buy cheaper tickets but they were still very expensive.
Is Horizon Zero Dawn perfectly suited to being transformed into a Game of Thrones video game? Just change the story line and convert all the machines into fantasy animals, and all the technology-based weapons into magic-based weapons.
For those interested in ADHD meds: I just learned of the existence of one called Guanfacine that's in a whole different drug class from the others, which all seem to be uppers. In fact sleepiness is the first side effect listed. I very quickly skimmed the research on its effectiveness, and my impression was that it was slightly less effective than adderall and the other familiar ones, but not much less. Seems likely to me that it's less abusable. Unlike the ADHD drugs in the uppers class it does not raise blood pressure -- in fact it lowers it some -- so it's a safer option for people with ADHD & high blood pressure. Has anyone here had any experience with it? What did you think of it?
Here’s a story about one of the ways GPT4 is dumb: I had asked google what household agents work to clean stained linoleum, and got hundreds of hits, most of which were exact duplicates of each other. The concoctions suggested were water + vinegar, water plus vinegar plus dish detergent, and water plus vinegar plus baking soda. The 3rd formula, though less popular than the first 2, showed up many many times. It’s obviously nonsense: The vinegar and baking soda will react to produce water, co2 & sodium acetate.
So I wanted to see how well ole Chat understood the Internet info about this topic. Asked it for household agents that might remove stains from linoleum. It said vinegar plus water plus dish soap, which I think was the most popular of the multiply-duplicated answers my google search turned up. So then I asked it whether including some baking soda would be a good idea, and it said “Adding baking soda to your cleaning mixture can enhance its effectiveness, especially for tackling tough stains and odors on linoleum floors. Baking soda is a mild abrasive, which makes it good for removing stuck-on messes without damaging the floor, and it's also a natural deodorizer.”
Yeech! Seems to me Chat’s process is essentially giving the answer that appears most frequently on the increasingly enshittified Internet, and not taking into account 3 things it surely knows: (1) There’s an online industry of copying advice and reposting it in a very slightly different format interlaced with ads. (In fact I asked Chat in a different conversation about the duplicates, and it was aware of the phenomenon and explained reasonably well why it exists.). (2) There are a lot of situations where how frequently a view is endorsed is not a measure of how likely it is to be true, and this is one of them. (3) Mixing vinegar and baking soda causes a chemical reaction in which each decomposes and in the process loses whatever stain-removing properties it had. (And Chat knows they react, too — even mentioned it in its advice: “Prepare the Solution: In a bucket, mix the hot water, white vinegar, and if using, a few drops of liquid dish soap. Gradually add the baking soda. Expect a mild reaction as the baking soda reacts with the vinegar, but this will subside quickly.”)
So it can’t put together different things it knows that would allow it to judge the online advice, just goes with the most popular. Makes me think about a post on here a few months ago in a discussion of whether some well known middle-aged white man, forget his name, was an appropriate appointee for some post — I believe the post was for the advisory board of an AI company. So the post somebody put up was Chat’s response to the question “Is Mr. M. A. Whiteman a good choice for this role?” And Chat gave a list of the guy's supposed good and bad qualities, as represented in the media, but came down more on the side of nay because more sources said Whiteman seemed to be impulsive. Chat’s summary seems to me just the linoleum stain recipe in another domain. It seems to be widely used for this purpose, and I think relying on it that way is going to make us all dumb way faster than embryo curation for Best of Breed can make us smarter.
And one more piece of snark: Chat said that adding baking soda to your cleaning solution can “enhance its effectiveness.” Nope. Even if it worked it would not enhance the effectiveness, it would increase it.
From *Catch 22*: “He was a spry, suave and very precise general who knew the circumference of the equator and always wrote "enhanced" when he meant "increased. He was a prick.” Chat’s a prick too.
I follow Peter Zeihan, who does geopolitical/economic prediction. I don't entirely trust him. Is he telling people what they want to hear? Still, he's interesting, and it's reasonable that demographics are pretty intractable. I haven't heard him say anything about the interactions between nations as population sizes change-- for example, big decline in China, much smaller decline in India, and what happens?-- I'm sure about China and fairly sure about India.
While we're on the subject, is it true that the really high end chips are only made in Taiwan?
Some of his predictions are 50 years out, and I'm willing to bet that Something Weird Will Happen Which Will Make Things Different.
I would count cheap gene sequencing and cheap computers as Big Weird Things. Recent plague and increase of war aren't as weird, but they did make a difference.
So here are some possible prediction-upsetting developments. Modest increase of longevity, enough to enable most people to be healthy until they're 90. Modest increase of intelligence-- say 20 IQ points for average people. Some element that was previously no big deal becomes crucial. A plague that's as infectious as measles. A new extremely popular religion.
I analyze the latest PISA results to figure out why Germany's performance has declined so much in recent years. My focus is on figuring out the extend to which changes in migration patterns can explain the decline. Remember to subscribe!
Introduction
In 2010, the book "Deutschland schafft sich ab" (Germany Abolishes Itself) was created by Thilo Sarrazin. This has made a lot of people very angry and has been widely regarded as a bad move. Sarrazin's core thesis on the topic of education can be roughly summarized as follows:
The German birth rate is low, with less than 1.4 children per woman. This is contrasted by a large number of migrants, especially from Muslim countries, who have higher birth rates.
Many migrants have educational deficits compared to the German population.
Even after several generations, these migrants do not catch up with German society. This is due to genetic and cultural inheritance as well as little pressure to integrate.
In the long run, Germany’s educational achievements will deteriorate due to this demographic change.
Sarrazin's critics argued that he was right about some things, but that he painted too bleak a picture and mixed truths with falsehoods. They pointed out, for example, that there had been progress in the area of education among Turks, a large Muslim immigrant group. Against the background of the recently published PISA study, in which Germany performed miserably, it seems appropriate to re-examine Sarrazin's thesis. In particular, I will use the latest PISA study to answer the question of whether, and to what extent, migration aspects play a role in the continuous decline of German education...
Some of you who are spaced-repetition nuts might remember a blog post series I wrote called Why Anki Works (https://nephewjonathan.substack.com/p/why-anki-works-part-1), which was the product of a year and a half of teaching Latin, trying to get the kids to use Anki, and failing--but learning a LOT about how and why spaced repetition works in the first place.
I'm now starting a startup--it doesn't have a name yet--for next-generation spaced repetition software, after spending a couple months getting my webdev skills up to speed on a personal project. There's a whitepaper here (https://eliotweb.net/srswhitepaper.pdf) describing what I'm building and why (some of the screenshots are out of date; section VI is also a bit more puffed-up than I'd like, since I was hoping to catch attention from some VCs, but I certainly don't disbelieve any of it). I'm currently working on the card designer/template maker (screenshot at https://eliotweb.net/templatemaker31124.png) and should have it more-or-less ready for the prototype by the end of this work week. Then it's on to card creation and reviews.
Big differences from Anki that have been implemented so far, most of which is visible in the screenshot, include a what-you-see-is-what-you-get card editor that uses markdown instead of HTML, fields that aren't just text/images (like checkboxes, buttons and dropdowns), the ability to make cards *contingent* on field values (so that you can have a template for German vocab create a card to test noun gender that's only created if the word is a noun), character replacement (so that you can type <scho"n>, <sophi'a>, <k'Atab>, or <gorod> and have it convert automatically to <schön>, <σοφία>, <כָּתַב>, or <город>), and the ability to set cards as prerequisites to each other, creating more advanced card relationships than Anki's sibling-burial. A schema for what this looks like can be found at https://eliotweb.net/reviewfrontier.png. Facts live in a network with each other, and a next-generation SRS platform should be able to mimic this. You *can* memorize the structure of an amino acid as a jumble of letters and lines, but it's parrot-knowledge if you don't know the functional groups.
Beyond that: the initial catalyst for working on this, as I said, was frustration with trying to get the kids to use it, so the long-term goal is to make it work for classrooms. That means it needs to live in the browser, rather than being installed locally, and although it should work well for autodidacts (who are likely to be the very first customers and have provided the initial spurt of funding), it should also have classroom abilities like Quizlet where teachers can make cards for students and see how they're doing (more details in the whitepaper). The knobs and dials mentioned in the last paragraph and seen on the screenshot are also motivated in large part by the realities of classroom usage. If you want to make custom card types in Anki, you'll need to know HTML (even for something as simple as a line break), and any halfway advanced auto-formatting requires Javascript--something most teachers don't want to waste time with, let alone their students. My own vocab cards have a massive Javascript block that color-codes the foreign-language field on the back side of cards based on their gender (so I can type <m> in the field and get blue coloring, or <f> and get red), but this is kludgy and non-obvious. So the general guideline here is that a 45-year-old Spanish or bio teacher, or ninth-grader, should be able to do pretty much anything an Anki power user can do without having to write any code. (And other things as well: it would be really nice to be able to create diagrams in-hourse, make clozes for those diagrams, and color-code MathJax formulas).
<Scott, feel free to remove this section if it shouldn't be included outside of a classified thread>
I talked to a couple of VCs, and there was some interest, but also a consensus that bootstrapping to a prototype via crowdfunding might work better early on. There's a GoFundMe for this purpose (https://www.gofundme.com/f/nextgeneration-spaced-repetition-software) --GoFundMe made me add a disclaimer to the bottom of the story saying I wouldn't be handing out any freebies, but I can say with a nod and a wink that backers will get very heavy discounts on subscriptions at launch (right now we're looking at a regular monthly individual subscription being $5, a yearly subscription $50, and a lifetime subscription $150, with backers' donations counting towards accounts at a third of the price--so if you donated $100 you'd get a lifetime subscription for yourself and a spare for a friend). Teacher accounts with group-management abilities will be a little bit pricier but come with a few free student accounts.
</possibly-inappropriate-section>
I've got a semi-cofounder--friend of a friend who just got laid off from his SWE job and is helping to work on this for a month or two to buffer his portfolio and get a small equity cut--but if you're an SRS nut who's got good webdev, backend or cybersecurity skills (especially the latter) and are interested in cofounding for the long haul, please drop me a line (either campbell · nilsen at gmail · com, or nephew_jonathan on Twitter--DMs are open). I don't have real funding yet (the GoFundMe is putting enough food on the table to not need to work full-time for a couple months while coding this), so until that happens, all I have to offer is equity. (Or, if you're just an SRS nut who's got an idea for something you think we should include--please, please drop me a line. Quite a few features have made their desirability known in the course of coding the prototype, and I'm sure there are others that escape me.)
I'm starting a new thing called Mindful Media Club. The aim is to trade tips on better personal use of social media - things like settings, browser plugins, social practices, to make it work for you and what you need and your mental health.
If you're in London, first meeting is on Wednesday March 20th at Newspeak House: https://lu.ma/7vouuhon
What's the relationship between the proofs for the existence of God and the evidence of God? Most Christians of my acquaintance say they believe because they have personal experience of God. That's excellent. I am happy for them. But then, who/what are the proofs for? If you already believe in God, why do you need proof?
And what does it mean to doubt the existence of God when you have evidence? Are there other domains where you have evidence of something but you doubt it anyway? Even St. Thomas required evidence before he believed in the risen Christ and I’m willing to bet that Thomas did not doubt again after poking the risen Christ’s wounds.
Hi all. I've come to plug my latest on viewing life as a series of group studies, each one spanning the full range of possible perspective. Also, Boyz II Men. Check it ✌️
Research questions: Liberal vs right wing morality (L’affaire Rittenhouse)
I’ve been looking into the question of where we can find fundamental moral disagreements between liberals, even right adjacent or sympathetic liberals and right wingers.
I think the Rittenhouse case provided the best possible example of one of these. Any number of anti-woke liberals will agree with right-wing observers on all the specific facts. Yet I haven’t yet found a single one who will with say, “Rittenhouse was not only legally not guilty of murder… but acted in a morally exemplary way and his actions that day were a positive good.” Nor for that matter will you get this response from any of the never Trumpers as far as I can tell. I think it represents one of the fundamental and unbridgeable gaps in american moral instincts.
Do you know of any anti-woke liberal who’ll state this (ideally a public figure, but personal anecdotes are fine)?
Can someone explain to me how simulation capture works? I recently read "The hour I first believed" and I'm confused as to why consciousness (as a mathematical object) would "split" when it gets simulated.
In this post I explore what core wrong assumption philosophers have been making while trying to solve it, why it's wrong, how comes so many managed not to figure it out for such a long time and what exactly the correct model looks like.
Are there any academic researchers or grad students who would like to pick up a little side gig here?
I am, and have been, curious what the academic literature has to say about "how to parent so that your adult children actually want to spend time around you". Not so shockingly this is for the personal reason of wanting my child to grow up to actually want to spend time around me. I'm particularly interested in variations between cultures and countries of origin.
I would love to hire someone to either do a quick survey, or to fund someone actually doing a deep dive and producing a literature review for publication, or something in between.
Has the fear of nuclear war dissipated somewhat, particularly amongst gen Z? Recently Macron said he didn’t rule out ground troops in Ukraine. That’s the kind of rhetoric that would frighten, and still does frighten Gen X or older. Yet, after he said it I noticed that Reddit was full of support. Nobody seems to realise the strength of the Russian nuclear armaments.
For a generation that is scared of the future, of solvable issues like climate change, this is a strange reaction.
Hello! A bit off topic, but I am running a anonymous 3-min salary survey in Berlin/Munich with a small team of volunteers. It's for Tech/Startups & corporate space. If you can participate or help me spread the word I would much appreciate:
I'm a possibly interesting guy doing possibly interesting things. Now in Egypt.
https://youtu.be/8_NW_yFC2Oc?feature=shared
https://ydydy.substack.com/
Six eighth-graders facing criminal charges for "hateful and racist" online chat: https://abcnews.go.com/US/eighth-graders-charged-racist-group-chat-southwick-massachusetts/story?id=108123039
But yeah, some rich affirmative action plagiarist losing her job at Harvard is the real authoritarianism
If you've noticed that Boeing has been in the news quite a bit recently, and not for the reasons they would like to be, you may be interested in this article.
https://backofmind.substack.com/p/how-the-wrong-side-won-at-boeing
It has some interesting things to say about the "Ricardian fallacy," and treating financial reports as the be-all-end-all of understanding the state of a business.
I remember reading recently... not sure if it was on Less Wrong or Astral Codex Ten... a great insight about colonialism: how it made the colonies poor, *without* necessarily making the empires rich. The colonies suffered a lot and had their wealth stolen, but the empires had to build armies which cost a lot of money and it is not obvious that the entire thing was a net benefit. So instead of modeling the situation as "X stealing money from Y", it would perhaps be better to see it as "X burning Y's money".
And I just realized how all this logic also perfectly applies to Soviet Union. Soviet Union made its satellite states poor *without* making itself rich. An average person in Russia was actually much poorer than e.g. an average person in Czechoslovakia. So from the perspective of the person in Czechoslovakia, it made sense to complain that their country was made poor by the decades of Soviet occupation... and from the perspective of the person in Russia, it made sense to object "what do you mean by 'poor'? look at me -- if you think that I have your money, you are crazy".
The answer is that there is no contradiction between "X making Y poor(er)" and "X not becoming richer in the process". Just like in real life, breaking someone else's legs doesn't make your legs healthier. But with wealth, people often have this fixed-sum intuition, and expect that if wealth was taken from somewhere, it *must* have appeared somewhere else (and if it did *not* appear somewhere else, then the person complaining about the theft is lying).
That makes me think about the later development of Eastern Europe. How much of the economical growth was the financial support from EU, and how much was... simply the Soviet Union not ruining things anymore. (This is not a dilemma; I assume that *both* numbers were huge.)
I’m not an expert at UK law, but I read that as two years in jail and an additional 2 years suspended, not the same 2 years suspended. If you and OP are right, that’s a valid point, but the more natural reading is that those are consecutive.
My project recently released this - it has been described as a return to OK Go style, true effort music videos - Inception meets Fear and Loathing, it imagines how democratised neuralink-like tech could unfold.
https://www.youtube.com/watch?v=W3d-cmd0GjU
It additionally plays with the metamodern tension between silliness and depth / irony and sincerity.
For fans of Techno-Buddhism, 80s Sci Fi with a touch of Tim and Eric.
Curious on interpretations of the message if anyone would like to deep dive it.
In the UK, you will *literally* receive a harsher penalty for distributing "racist" stickers than for RAPING a child (well, at least if you're a brown rapist, not sure about if you're white).
https://www.bbc.com/news/uk-england-leeds-68448867
https://www.bbc.com/news/uk-england-tyne-68446855
Why can't we be more like enlightened Europe?
I'm trying to find a specific SSC post from years ago. In it, I remember that Scott defines what "common knowledge" is, and something about how it involves an infinite regression of that everyone knows that everyone else knows that everyone else knows something is true. I don't think it was It Was You That Made My Eyes Blue, I think it came earlier than that. I remember maybe something about a teacher asking kids to stand up in a classroom...? Any help appreciated. Thanks!
As someone who initially thought that Rittenhouse was probably guilty but eventually came around to it being legitimate self defence, I do think that I keep an open mind. And thus it's important to highlight when I have a major update in the *other* direction as well. I'd never paid much attention to the Floyd case before and had gotten an inaccurate picture due to hanging out in conservative circles.
Last night, I read this post about this case, and the details are far, far worse than I'd imagined. It really is appalling. And thus I'm reposting this in the hopes that everyone else reads it as well.
https://radleybalko.substack.com/p/the-retconning-of-george-floyd
So, if this article is to be taken seriously, Google has become a DEI basket case.
https://www.piratewires.com/p/google-culture-of-fear
Anyone know how trustworthy Pirate Wires or Mike Solana are?
Does anyone else worry about the concept of online "bait", and how it's becoming more common to ignore anyone with a wildly different view or opinion?
Any idea that's sufficiently idiotic or ill-informed seems to be immediately written off as 'someone trying to mess with people for a reaction', and deemed not worthy of your time. But it seems perfectly plausible that there could be someone unironically touting such views, and in that case, do we not just reinforce that person's echo chamber by choosing not to debate them? Where do you draw the line?
I ask because I see plenty of obvious bait on a daily basis - that's not the stuff I'm referring to. Rather, I mentioned a post in this very comment section, and my partner immediately said "ha, and people were actually replying to them?" like it was an impressive feat of trickery, but the vibe I got was way more sincere. If we start treating every "crazy" opinion we see as "bait", then are we not just siloing the people who honestly hold these views?
Forget China, Japan, and Europe. I just discovered that *Uzbekistan* of all places has a bullet train. And they built it in 2011, in only seven months, less time than it takes an American builder just to scratch their ass. Now I'm really feeling civilizational inadequacy.
For those who like mixing good deeds an capitalism: there is a hole in our financial system. We have banks to fund mundane businesses. We have Wall St. to fund huge corporations. We have Venture Capital to fund businesses which have the potential to become huge in a reasonably short period of time, or be sellable to a huge venture.
But what of creative businesses which have moderate, but nontrivial risk, but don't have the potential to become Wall St. sized? We need an equity market for such businesses so VCs and angels can cash out. Here is a proposal for such a market:
https://conntects.net/blogPosts/HolisticPolitics/137
Does anyone have any recommendations for "political" novels? I don't mean defending any particular political system or policy. I more mean a novel of factions and scheming and plotting and stuff. It doesn't even have to be in the arena of actual government. I'm also super genre-agnostic - though I've been on a sci-fi kick lately.
On last week's open thread I replied to a request by user Nancy Lebovitz, asking libertarian readers of this substack what laws or parts of government [they would] like to take down first"(https://open.substack.com/pub/astralcodexten/p/open-thread-318?r=7vbt1&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=50882478).
Among other things (https://open.substack.com/pub/astralcodexten/p/open-thread-318?r=7vbt1&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=50884197), I advocated for lifting the statute of limitations on property crimes and partially doing away with the proscription on applying the law retroactively (ex post facto law).
Basically, I suggested that something like the Radbruch formula be applied to historical laws backwards pretty much indefinitely, rectifying all unjust transfer of property as much as possible.
Specifically, I suggested the following rough test:
"a) Was the transfer of property in question legal at the time?
b) If so, were the institutions of the time that governed the transfer morally or legally uncontested at the time for reasons widely accepted today?
c) If so, can these institutions reasonably be considered just when viewed in light of fundamental rights that ought to hold at any time or place?
d) If the transfer in question fails tests a,b or c, can the plaintiff prove a causal connection from the unjust transfer of property to present outcomes?
If so there should be reparations, regardless of how much time has passed since the event."
This being a rather radical view, my proposal drew a fair amount of criticism and ridicule. This came mostly in the form of putative counterexamples which my approach would not be able to handle or in which it would yield absurd consequences.
In response to Deiseach, I promised to rebut these with my own worked examples (https://open.substack.com/pub/astralcodexten/p/open-thread-318?r=7vbt1&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=51031434).
So without further ado, here we go:
Example 1 (thanks to Arrk Mindmaster): "Nanni and Ea-Nasir"
Arrk Mindmaster brough up this example as a reductio ad absurdum (https://open.substack.com/pub/astralcodexten/p/open-thread-318?r=7vbt1&utm_campaign=comment-list-share-cta&utm_medium=web&comments=true&commentId=50963605), so let me address it (details of the case here: https://en.wikipedia.org/wiki/Complaint_tablet_to_Ea-n%C4%81%E1%B9%A3ir). Their challenge consisted in 1) the absurdity of such a case succeding in court and 2) the absurdity of an immense award due to compound interest. I shall address both issues.
Suppose we found out that Ea-Nasir indeed never refunded Nanni his money and some descendant of Nanni would bring a case against some descendant of Ea-Nasir today. Suppose further that Ea-Nasir indeed defrauded Nanni according to the laws of the time failing part a) of the test. Finally, assume that the standard of proof to be met is "preponderance of the evidence" which is usually required in civil cases.
Reply to 1) Plaintiff's lawsuit would fail because they would almost certainly fall short of meeting part d) of the test, which requires proving causality.
This is for several reasons:
i. Dilution: if Ea-Nasir has any living descendants at all, then approximately everyone on earth is his descendant (https://www.theguardian.com/science/commentisfree/2015/may/24/business-genetic-ancestry-charlemagne-adam-rutherford). The same goes for Nanni. So approximately everyone could appear on both sides of the lawsuit and the amounts owed would be infinetisemal. Any possible causal influence has been diluted and spread around among both parties descendants so much that it would almost certainly be a wash.
ii. Non-ergodicity of economic growth: while global GDP has increased manyfold since ancient Babylonian times, this does not mean that individual fortunes have. Time average and ensemble average of the wealth building stochastical process diverge. In particular, fortunes decay over time (and symmetrically do debts): conventional wisdom says that fortunes are built in one generation and lost by the third generation. While this belief has been challenged (https://cdn.ymaws.com/my.ffi.org/resource/resmgr/docs/goodman_study.pdf; although I will in a later post present some evidence that the three generations rule may not be far off after all), it remains true that fortunes that are more than a few hundred years old are exceedingly rare. This is surely not unrelated to the fact that debtors' heirs can usually simply refuse to accept the inheritance.
iii. Lack of data: obviously.
In light of this, any suit being brought that is based on claims older than a few hundred years is almost certainly going to fail. I would expect that after a change of the law a few such lawsuits would be attempted largely for publicity but then rather quickly fizzle out.
Reply to 2) The value of ancient claims would not skyrocket to astronomical heights. Insofar as one can establish causality, one would also likely get an estimate of the effect size. For the reasons already stated, it is not likely to have increased dramatically through the generations but instead would probably have decreased. For this reason a fixed rate of interest is simply completely unsuitable to simulate the development of the value of a claim over long time periods. The interest rate is meant to reflect the counterfactual development of the value of the property of which the plaintiff was illegally or unjustly deprived. But empirically speaking people don't put their money into a savings-account, ETF or similar over many generations. Rather they consume it, make bad investment decisions or just get unlucky. Moreover, until a few hundred years ago GDP-per-capita was essentially flat (https://ourworldindata.org/grapher/global-average-gdp-per-capita-over-the-long-run). In such time one could not expect one's assets to appreciate over time even if carefully managed. For these reasons, courts would very likely not use a fixed interest rate to determine the present value of a claim but would rather estimate it from the causal evidence that the plaintiff has provided.
This was only the first worked example, but it is getting late here. Tomorrow I will try to post one on slavery in the US. That one will in particular include some concrete values.
Anytime I get through reading a Links post (not here necessarily, just any listing pointing to lots of Substacks), my primary thought is "man, it annoys me how many blogs won't let you Like a post without first subscribing". Doesn't matter if it's a free tier sub, Beware Trivial Inconveniences. Turning off comments to non-subs makes sense for the same reason, but...what, it's important to prevent frivolous drive-by likes from the uncommitted? I guess? Seems like a bad deal for new blags hurting for exposure, which tends to be what ends up on such roundups in the first place.
Semi-relatedly, I swear that sometimes Substack loses track of historical likes. Will occasionally reread something I'm certain I liked (both senses) a year ago or whatever, and that heart is just...gone. Which is frustrating, since an existing heart serves as a good bookmark that yes, I did in fact read this before, and can check it off the list after a brief skim.
I have two new posts on monetary policy:
https://thomaslhutcheson.substack.com/p/framework-for-monetary-policy-1
https://thomaslhutcheson.substack.com/p/framework-for-monetary-policy-2
and one on macroeconomic data improvements
https://thomaslhutcheson.substack.com/p/improvements-in-macroeconomic-data
I'm curious if anyone in the community works in the GIS/Geospatial field. I'm about to write my bachelor's thesis in Urban Planning, and have become interested in a career in GIS. I've browsed https://80000hours.org/ for GIS work that might have a significant positive impact, but come up short.
I'd be grateful if anyone would like to chat/give some advice on three main questions:
1. Whether a career in GIS/Geospatial is worth pursuing.
2. Potential research topics for my bachelor's thesis.
3. Whether to pursue a master's degree or start looking for work after I've graduated.
Thanks!
How is the accelerationist view coming? We went from ChatGTP 3.5 (2022-11) to ChatGTP 4 (2023-03) in four months, but now twelve more months have passed and I think it's fair to say that there has basically been 0 progress in text-based AI since then? No GTP5, lots of new competing models but none that are widely believed to be superior to GTP4. Sora was touted as a game-changer for video AI when it was announced, but it has no official release date and we have little info on its average quality or cost to run.
I realize this comment could age like milk any second, but is the current evidence consistent with a view that things are plateauing or at best advancing linearly rather than exponentially? One response could be that it's zooming in too far to compare the last few months to the previous few months, and indeed, the next few months could certainly show another spurt of progress. But on the other hand, isn't the concept of accelerationism / exponential growth that the curve is the same no matter how you zoom in, and progress durations are supposed to get shorter and shorter, not longer and/or the same?
Freddie deBoer wrote a recent-ish article on Israel: This is Zion, https://freddiedeboer.substack.com/p/this-is-zion.
The article mainly argues that the USA is "better at being Israel than Israel", in the words of the writer. That is, the USA's population of Jews is 85% of Israel's, and they have a better quality of life across a broad spectrum of metrics than their Israeli counterparts. The USA also absorbs Jewish immigrants with less hostility and less friction than other groups, including - first and foremost - Arabs and/or Palestinians and/or Muslims.
Article aside, this section got my attention:
>> Defenders of the modern Israeli state are in this constant argumentative bind: they must ceaselessly insist that Israel is teetering on the brink of destruction, in order to keep American money and weapons and diplomatic muscle flowing, while at the same time claiming that Israel is the only place where Jews can be safe. These are, obviously, directly contradictory sentiments. If it takes the constant patronage of the most powerful nation on earth to keep Israel from destruction, and even then the country is subject to assaults like that of October 7th, in what sense could Israel possibly be considered a safe place for Jews? Well, you could point out that by most basic statistics that catalog violent threats, Israelis are safer than citizens of most countries in the world. (If you ask an Israeli whether their country is safer than Chad or Colombia or Pakistan, they’ll get offended that you asked.) Unfortunately, you are then merely pulled back into the other side of the paradox - if it’s true that Israelis are remarkably physically safe, in context with much of the rest of the world, how can we justify the seemingly perpetual outlay of vast amounts of American ordnance and treasure on Israel’s behalf?
I think this is a general pattern in politics and similar ingroup-outgroup dynamics, not specific to Pro-Israel supporters, people simultaneously want to believe "The Enemy is Strong" and "We Are Strong". They selectively switch between those 2 according to which is most convenient at the moment, I have caught myself doing it several times in other issues not related to Israel-Palestine.
Anyway, the article shines a light on the many contradictions at the heart of Zionism: Israel is the *holy* land, Israel is a *secular* democracy. Israel is the *safe haven* for Jews everywhere, Israel is in *existential danger* and you can't hold a state to high moral standards when its existence is at the line. Most moral army in the universe, but don't hold it to higher standards than any other army, that's antisemitism. Israel is the only place where Jews can truly be safe, Jews in Israel are terribly unsafe (and the Palestinians are at fault).
Scott mentioned the problems with substack loading recently (much more of a problem on mobile for me). It seems like this is almost definitely due to the comments sections on posts all loading with the page, something that Scott's blog does differently from others, I believe because of Scott's requests. I know Scott doesn't want likes on comments so you can't just display the top comments, but is there some other solution, like only displaying the first 20 top level comments and then have a See More button after that?
I believe I remember Scott wanting to have the comments here replicate the way they worked on SlateStarCodex, but it seems like that's causing technical problems at this size of community.
On the morality of shooting Arsonists (And why we hate liberals):
A community near you sees two nights of it's businesses being burned down in the name of a political cause. In each case the extremely well funded government security force stands by indifferently while the very narratives which justified the attacks are propagated by government and press.
On the third day you show up armed to the perimeter of one of the few businesses that hasn't been destroyed. You like this business. It sold you your first car. It's a large business, a car lot so there's no possible way you can block all of it. All the cars in it are close together, so lighting up one means eventually lighting up all. A mob gathers and a man with a massive tank of gasoline and a zippo lighter screams, 'we are gonna burn this place down, nothing will happen to you.' You stand your ground and point your gun at him. He backs off, and runs. You believe he is retreating, but then you realize he's running towards another part of the massive parking lot. He pours the gas on a car, a prepares to throw the zippo lighter. You raise your gun, shoot him straight through the chest and kill him.
Was this morally justified?
---
The average natural right winger instinctively says yes. If you ask them why, most of them couldn't give you a great answer. But I'll give you a few now, each of which I've heard from relatively unintelligent people (if stated less eloquently):
- Property is the blood, sweat, tears and years of human life spent building it (ok normally not the blood part). To destroy it, is in a very real sense to kill a part of a person.
- You know better than anyone else what the value of your life is and if you think it's worth less than the pleasure you derive from destruction, I'm not gonna deny your assessment.
- I agree not to kill in exchange for the government agreeing to protect basic rights. If the government blatantly refuses to do so, only the laws of war and nature apply.
- A random business is not a legitimate target in a political conflict, destroying it has no value to whatever war the rioters are engaged in. They are war criminals.
(Of course, all of the above combined is it's own argument.)
I'm close to sure that there is at most a single digit percentage of educated liberals who take this position. I know this because I've gone over the public statements of any number of anti-woke liberals on the Rittenhouse case (a case infinitely more defensible than mine), and found no one who is willing to state unambiguously 'he was right to do what he did, including being there in the first place.'
Man in the liberal vision, is but property of his government or society; and if someone decides to destroy decades of a life's work and no one 'authorized by law' is coming to save him - well, it sucks for him. After all, it might, in some final utilitarian accounting be bad for society if a man took the law into his own hands - even in response to society's total abandonment of it. Of course, in any right wing vision, at least the non-religious ones a man who accepted this is no man at all.
Which is why, when push comes to shove; no compromise between liberals and right-wingers is possible. All talk of 'more in common' is empty. We believe that you would impose upon us norms that we could only wish on our worst enemies. That you might not personally hate us is of no consequence as we believe that you are fundamentally incapable of love for anyone at all. For all our enmity towards Communists and/or the woke, we can at least imagine a world from their perspective in which business owners are the real enemies who must be crushed to create a classless society. And boy, are they dedicated to this cause. Anyone has to admire that in the same way Orwell admired Hitler.
But for you, you who would condemn a man because he wasn't carrying a badge, because 'life is worth more than property', we have nothing but hatred.
Has anyone else noticed Google Doc's autocorrect get much more aggressive and stupider recently?
Previously, it would just attempt to correct mispellings or grammar errors. Sometimes it would get one wrong, but for the most part it was ok, and it would never attempt to apply the same correction more than once if you backspaced and changed it back. However, perhaps a few weeks ago, it got a lot more aggressive and also stupider, and now frequently makes changes that aren't even gramatically correct.
For example, when I was writing just now, in the space of one paragraph, it changed the "worn" to "worms" in "However, there were also small holes worn", and it changed "I applied" to "I apple".
Furthermore, it no longer gives up if you undo its correction. I've seen it apply the exact same (mis)correction three or more times when I backspace and fix the "correction".
Does anyone here know what's going on here? I'm curious what people on Google Docs would say about this.
I've been learning German for almost 2 years now. I've used a combination of in person classes, remote classes, and Duolingo. Across all of them, I probably average an hour-ish per day across those two years.
One thing I've noticed about Duolingo that I wish would be improved is that, while it occasionally will bring you old sentences with vocab that you haven't seen before, it's A) the same set of sentences you used when you initially used the word and B) still sentences where the entire context of the sentence is designed to help you figure out what the word is.
B is extremely helpful when you are first learning a word, but, in my opinion, it becomes detrimental when you are trying to cement a word that you already know, and prevents you from putting that word in a larger context.
I realize that generating a wide range of novel sentences for every vocab word, especially when those sentences would ideally have a decreasing gradient of contextual information, would be prohibitively expensive. This feels like the kind of problem that an LLM would be perfect for solving.
Language learning feels like the area of learning/education that is _most_ ripe for being completely upended by LLM assistants. I know that Duolingo is slowly trialing their "Max" AI assistant, but it doesn't seem like they are using it in an optimal way.
Duolingo is, for me, currently just a useful side thing that helps support and reinforce the learning I do in my classes. I think with a well thought out and intelligently implemented LLM assistant, it could become my primary learning tool. Unfortunately, I'm not sure that that would jive with their current incentives which lean towards gamification and just getting people to open the app every day. So Duolingo may not be the thing that does it, but man it feels like _someone_ is going to make the thing I'm envisioning, and soon.
For those who speak Chinese: I am looking for a name for a character in a story I'm writing. It does not need to be a word typically used as a name. It should be a term meaning something like "clever and brave." It should have a clearly positive connotation -- so it should not be commonly used to describe, say, talented criminals who are sneaky and sly. I asked on the Reddit Chinese language sub and got zero answers. Wow, Reddit's a lot less friendly these days! I'd be very grateful for suggestions here.
I'm trying to decide, Scott, whether you are the worst employment-opportunity signal booster in the world or the best....I will of course have a 1,000-word comment exploring this question ready soon.
(Which, in my ongoing effort to remain a sane person, I will then delete unposted.)
So as spring comes early yet again here in Western NY, I'm reminded of an observation*. Which is that global warming seems to be mostly a warming of the coldest temperatures and not an increase in the hottest temperatures. So in the northeast, winters are less cold and shorter, and the summers are longer, but don't really get any hotter... only in that there are more days with high temperatures.
*I read this in "Unsettled" by S. Koonin. See figure 5.5 here. https://d2fahduf2624mg.cloudfront.net/pre_purchase_docs/BK_COMM_006119/2021-05-05-05-04-59/bk_comm_006119.pdf
I'll be honest, I laughed out lout at the EA signal boost
There's been something pretty interesting going in philosophy blog Bentham's Bulldog: the utilitarian author has started to believe in God because it makes more sense than atheism. He doesn't hold theism with 100% certainty, but he sees it as more likely than atheism.
He's written a lot on how he came by this view, but this is a good summary: https://benthams.substack.com/p/a-crisis-of-faithlessness
Does anyone here buy & sell individual stocks (or bonds, or crypto, or futures, or options, or commodities, or anything else?) I.e. does anyone here day trade, swing trade, or take longterm position trades? Anything other than just buying boring index funds and taking the market average. No judgement if you do, just totally curious- what motivates people to 'trade'? Are you profitable? More profitable than if you'd just held a boring index fund? Including taxes, which should take a significant chunk of any profits.
I spent a lot of time reading all the classic works on value investing, and I came to the conclusion that it's not worth the time investment for me to try to buy individual stocks and beat the market. But just interested in hearing peoples' experiences, good or bad
Hello!
Anyone know exactly what the point of Austin's SXSW is? They have setious ppl speaking but also Meghan Markle speaking, they once had Barack Obama speaking, and I think, music concerts. You could listen or read these speakers' ideas elsewhere like on YouTube. I'm curious why anyone pays $2k to attend. I tried to find ways to buy cheaper tickets but they were still very expensive.
Is Horizon Zero Dawn perfectly suited to being transformed into a Game of Thrones video game? Just change the story line and convert all the machines into fantasy animals, and all the technology-based weapons into magic-based weapons.
Does having $2 million make you a "multimillionaire"?
For those interested in ADHD meds: I just learned of the existence of one called Guanfacine that's in a whole different drug class from the others, which all seem to be uppers. In fact sleepiness is the first side effect listed. I very quickly skimmed the research on its effectiveness, and my impression was that it was slightly less effective than adderall and the other familiar ones, but not much less. Seems likely to me that it's less abusable. Unlike the ADHD drugs in the uppers class it does not raise blood pressure -- in fact it lowers it some -- so it's a safer option for people with ADHD & high blood pressure. Has anyone here had any experience with it? What did you think of it?
Here’s a story about one of the ways GPT4 is dumb: I had asked google what household agents work to clean stained linoleum, and got hundreds of hits, most of which were exact duplicates of each other. The concoctions suggested were water + vinegar, water plus vinegar plus dish detergent, and water plus vinegar plus baking soda. The 3rd formula, though less popular than the first 2, showed up many many times. It’s obviously nonsense: The vinegar and baking soda will react to produce water, co2 & sodium acetate.
So I wanted to see how well ole Chat understood the Internet info about this topic. Asked it for household agents that might remove stains from linoleum. It said vinegar plus water plus dish soap, which I think was the most popular of the multiply-duplicated answers my google search turned up. So then I asked it whether including some baking soda would be a good idea, and it said “Adding baking soda to your cleaning mixture can enhance its effectiveness, especially for tackling tough stains and odors on linoleum floors. Baking soda is a mild abrasive, which makes it good for removing stuck-on messes without damaging the floor, and it's also a natural deodorizer.”
Yeech! Seems to me Chat’s process is essentially giving the answer that appears most frequently on the increasingly enshittified Internet, and not taking into account 3 things it surely knows: (1) There’s an online industry of copying advice and reposting it in a very slightly different format interlaced with ads. (In fact I asked Chat in a different conversation about the duplicates, and it was aware of the phenomenon and explained reasonably well why it exists.). (2) There are a lot of situations where how frequently a view is endorsed is not a measure of how likely it is to be true, and this is one of them. (3) Mixing vinegar and baking soda causes a chemical reaction in which each decomposes and in the process loses whatever stain-removing properties it had. (And Chat knows they react, too — even mentioned it in its advice: “Prepare the Solution: In a bucket, mix the hot water, white vinegar, and if using, a few drops of liquid dish soap. Gradually add the baking soda. Expect a mild reaction as the baking soda reacts with the vinegar, but this will subside quickly.”)
So it can’t put together different things it knows that would allow it to judge the online advice, just goes with the most popular. Makes me think about a post on here a few months ago in a discussion of whether some well known middle-aged white man, forget his name, was an appropriate appointee for some post — I believe the post was for the advisory board of an AI company. So the post somebody put up was Chat’s response to the question “Is Mr. M. A. Whiteman a good choice for this role?” And Chat gave a list of the guy's supposed good and bad qualities, as represented in the media, but came down more on the side of nay because more sources said Whiteman seemed to be impulsive. Chat’s summary seems to me just the linoleum stain recipe in another domain. It seems to be widely used for this purpose, and I think relying on it that way is going to make us all dumb way faster than embryo curation for Best of Breed can make us smarter.
And one more piece of snark: Chat said that adding baking soda to your cleaning solution can “enhance its effectiveness.” Nope. Even if it worked it would not enhance the effectiveness, it would increase it.
From *Catch 22*: “He was a spry, suave and very precise general who knew the circumference of the equator and always wrote "enhanced" when he meant "increased. He was a prick.” Chat’s a prick too.
I follow Peter Zeihan, who does geopolitical/economic prediction. I don't entirely trust him. Is he telling people what they want to hear? Still, he's interesting, and it's reasonable that demographics are pretty intractable. I haven't heard him say anything about the interactions between nations as population sizes change-- for example, big decline in China, much smaller decline in India, and what happens?-- I'm sure about China and fairly sure about India.
While we're on the subject, is it true that the really high end chips are only made in Taiwan?
Some of his predictions are 50 years out, and I'm willing to bet that Something Weird Will Happen Which Will Make Things Different.
I would count cheap gene sequencing and cheap computers as Big Weird Things. Recent plague and increase of war aren't as weird, but they did make a difference.
So here are some possible prediction-upsetting developments. Modest increase of longevity, enough to enable most people to be healthy until they're 90. Modest increase of intelligence-- say 20 IQ points for average people. Some element that was previously no big deal becomes crucial. A plague that's as infectious as measles. A new extremely popular religion.
Suggestions for other weird possibilities?
Does Germany abolish itself? https://grauwacht.substack.com/p/does-germany-abolish-itself
Schafft Deutschland sich ab? https://grauwacht.substack.com/p/schafft-deutschland-sich-ab
I analyze the latest PISA results to figure out why Germany's performance has declined so much in recent years. My focus is on figuring out the extend to which changes in migration patterns can explain the decline. Remember to subscribe!
Introduction
In 2010, the book "Deutschland schafft sich ab" (Germany Abolishes Itself) was created by Thilo Sarrazin. This has made a lot of people very angry and has been widely regarded as a bad move. Sarrazin's core thesis on the topic of education can be roughly summarized as follows:
The German birth rate is low, with less than 1.4 children per woman. This is contrasted by a large number of migrants, especially from Muslim countries, who have higher birth rates.
Many migrants have educational deficits compared to the German population.
Even after several generations, these migrants do not catch up with German society. This is due to genetic and cultural inheritance as well as little pressure to integrate.
In the long run, Germany’s educational achievements will deteriorate due to this demographic change.
Sarrazin's critics argued that he was right about some things, but that he painted too bleak a picture and mixed truths with falsehoods. They pointed out, for example, that there had been progress in the area of education among Turks, a large Muslim immigrant group. Against the background of the recently published PISA study, in which Germany performed miserably, it seems appropriate to re-examine Sarrazin's thesis. In particular, I will use the latest PISA study to answer the question of whether, and to what extent, migration aspects play a role in the continuous decline of German education...
Read more: https://grauwacht.substack.com/p/does-germany-abolish-itself
Some of you who are spaced-repetition nuts might remember a blog post series I wrote called Why Anki Works (https://nephewjonathan.substack.com/p/why-anki-works-part-1), which was the product of a year and a half of teaching Latin, trying to get the kids to use Anki, and failing--but learning a LOT about how and why spaced repetition works in the first place.
I'm now starting a startup--it doesn't have a name yet--for next-generation spaced repetition software, after spending a couple months getting my webdev skills up to speed on a personal project. There's a whitepaper here (https://eliotweb.net/srswhitepaper.pdf) describing what I'm building and why (some of the screenshots are out of date; section VI is also a bit more puffed-up than I'd like, since I was hoping to catch attention from some VCs, but I certainly don't disbelieve any of it). I'm currently working on the card designer/template maker (screenshot at https://eliotweb.net/templatemaker31124.png) and should have it more-or-less ready for the prototype by the end of this work week. Then it's on to card creation and reviews.
Big differences from Anki that have been implemented so far, most of which is visible in the screenshot, include a what-you-see-is-what-you-get card editor that uses markdown instead of HTML, fields that aren't just text/images (like checkboxes, buttons and dropdowns), the ability to make cards *contingent* on field values (so that you can have a template for German vocab create a card to test noun gender that's only created if the word is a noun), character replacement (so that you can type <scho"n>, <sophi'a>, <k'Atab>, or <gorod> and have it convert automatically to <schön>, <σοφία>, <כָּתַב>, or <город>), and the ability to set cards as prerequisites to each other, creating more advanced card relationships than Anki's sibling-burial. A schema for what this looks like can be found at https://eliotweb.net/reviewfrontier.png. Facts live in a network with each other, and a next-generation SRS platform should be able to mimic this. You *can* memorize the structure of an amino acid as a jumble of letters and lines, but it's parrot-knowledge if you don't know the functional groups.
Beyond that: the initial catalyst for working on this, as I said, was frustration with trying to get the kids to use it, so the long-term goal is to make it work for classrooms. That means it needs to live in the browser, rather than being installed locally, and although it should work well for autodidacts (who are likely to be the very first customers and have provided the initial spurt of funding), it should also have classroom abilities like Quizlet where teachers can make cards for students and see how they're doing (more details in the whitepaper). The knobs and dials mentioned in the last paragraph and seen on the screenshot are also motivated in large part by the realities of classroom usage. If you want to make custom card types in Anki, you'll need to know HTML (even for something as simple as a line break), and any halfway advanced auto-formatting requires Javascript--something most teachers don't want to waste time with, let alone their students. My own vocab cards have a massive Javascript block that color-codes the foreign-language field on the back side of cards based on their gender (so I can type <m> in the field and get blue coloring, or <f> and get red), but this is kludgy and non-obvious. So the general guideline here is that a 45-year-old Spanish or bio teacher, or ninth-grader, should be able to do pretty much anything an Anki power user can do without having to write any code. (And other things as well: it would be really nice to be able to create diagrams in-hourse, make clozes for those diagrams, and color-code MathJax formulas).
<Scott, feel free to remove this section if it shouldn't be included outside of a classified thread>
I talked to a couple of VCs, and there was some interest, but also a consensus that bootstrapping to a prototype via crowdfunding might work better early on. There's a GoFundMe for this purpose (https://www.gofundme.com/f/nextgeneration-spaced-repetition-software) --GoFundMe made me add a disclaimer to the bottom of the story saying I wouldn't be handing out any freebies, but I can say with a nod and a wink that backers will get very heavy discounts on subscriptions at launch (right now we're looking at a regular monthly individual subscription being $5, a yearly subscription $50, and a lifetime subscription $150, with backers' donations counting towards accounts at a third of the price--so if you donated $100 you'd get a lifetime subscription for yourself and a spare for a friend). Teacher accounts with group-management abilities will be a little bit pricier but come with a few free student accounts.
</possibly-inappropriate-section>
I've got a semi-cofounder--friend of a friend who just got laid off from his SWE job and is helping to work on this for a month or two to buffer his portfolio and get a small equity cut--but if you're an SRS nut who's got good webdev, backend or cybersecurity skills (especially the latter) and are interested in cofounding for the long haul, please drop me a line (either campbell · nilsen at gmail · com, or nephew_jonathan on Twitter--DMs are open). I don't have real funding yet (the GoFundMe is putting enough food on the table to not need to work full-time for a couple months while coding this), so until that happens, all I have to offer is equity. (Or, if you're just an SRS nut who's got an idea for something you think we should include--please, please drop me a line. Quite a few features have made their desirability known in the course of coding the prototype, and I'm sure there are others that escape me.)
I'm starting a new thing called Mindful Media Club. The aim is to trade tips on better personal use of social media - things like settings, browser plugins, social practices, to make it work for you and what you need and your mental health.
If you're in London, first meeting is on Wednesday March 20th at Newspeak House: https://lu.ma/7vouuhon
And there's a Discord for anyone everywhere: https://discord.gg/hbSeb78cCv
What's the relationship between the proofs for the existence of God and the evidence of God? Most Christians of my acquaintance say they believe because they have personal experience of God. That's excellent. I am happy for them. But then, who/what are the proofs for? If you already believe in God, why do you need proof?
And what does it mean to doubt the existence of God when you have evidence? Are there other domains where you have evidence of something but you doubt it anyway? Even St. Thomas required evidence before he believed in the risen Christ and I’m willing to bet that Thomas did not doubt again after poking the risen Christ’s wounds.
More wondering here:
https://raggedclown.substack.com/p/arguments-for-god
Hi all. I've come to plug my latest on viewing life as a series of group studies, each one spanning the full range of possible perspective. Also, Boyz II Men. Check it ✌️
https://kyleimes.substack.com/p/there-is-a-cycle-that-never-stops
Research questions: Liberal vs right wing morality (L’affaire Rittenhouse)
I’ve been looking into the question of where we can find fundamental moral disagreements between liberals, even right adjacent or sympathetic liberals and right wingers.
I think the Rittenhouse case provided the best possible example of one of these. Any number of anti-woke liberals will agree with right-wing observers on all the specific facts. Yet I haven’t yet found a single one who will with say, “Rittenhouse was not only legally not guilty of murder… but acted in a morally exemplary way and his actions that day were a positive good.” Nor for that matter will you get this response from any of the never Trumpers as far as I can tell. I think it represents one of the fundamental and unbridgeable gaps in american moral instincts.
Do you know of any anti-woke liberal who’ll state this (ideally a public figure, but personal anecdotes are fine)?
Can someone explain to me how simulation capture works? I recently read "The hour I first believed" and I'm confused as to why consciousness (as a mathematical object) would "split" when it gets simulated.
I've recently solved the Sleeping Beauty problem. https://en.wikipedia.org/wiki/Sleeping_Beauty_problem
In this post I explore what core wrong assumption philosophers have been making while trying to solve it, why it's wrong, how comes so many managed not to figure it out for such a long time and what exactly the correct model looks like.
https://www.lesswrong.com/posts/gwfgFwrrYnDpcF4JP/the-solution-to-sleeping-beauty
Are there any academic researchers or grad students who would like to pick up a little side gig here?
I am, and have been, curious what the academic literature has to say about "how to parent so that your adult children actually want to spend time around you". Not so shockingly this is for the personal reason of wanting my child to grow up to actually want to spend time around me. I'm particularly interested in variations between cultures and countries of origin.
I would love to hire someone to either do a quick survey, or to fund someone actually doing a deep dive and producing a literature review for publication, or something in between.
I feel like I have the potential to be the worst candidate for the Effective Altruism position, which feels appropriate to 2024. Should I apply?
Has the fear of nuclear war dissipated somewhat, particularly amongst gen Z? Recently Macron said he didn’t rule out ground troops in Ukraine. That’s the kind of rhetoric that would frighten, and still does frighten Gen X or older. Yet, after he said it I noticed that Reddit was full of support. Nobody seems to realise the strength of the Russian nuclear armaments.
For a generation that is scared of the future, of solvable issues like climate change, this is a strange reaction.
Hello! A bit off topic, but I am running a anonymous 3-min salary survey in Berlin/Munich with a small team of volunteers. It's for Tech/Startups & corporate space. If you can participate or help me spread the word I would much appreciate:
https://forms.gle/wwLfMyaFS9rwNdVaA