I've had a scammer try to lead me through a cashiers check scam. (Briefly: I was selling an item on Craigslist, they said they'd mail me a check and could I pay the movers out of it. The scam continues with "the check is fake but it's a cashier's check so the bank gives you the money now, then someone -- likely you -- has to pony up when it's revealed the check is fake after all".) Not knowing the scam beforehand, I played along one step farther than I would've liked, and am currently in possession of the (presumed fake) check.
Question 1: is there some way to punish this behavior? I haven't been materially harmed (yet; the scammer does know an uncomfortable amount of information about me), but presumably writing fake checks is illegal even if they're not cashed. But, who would I report this to? My police department? "Their" police department (based on the address)?
Question 2: is there even enough information for anyone to do anything about it? I have a cashier's check (allegedly from a small business, not a person), a letter by which it arrived (which has a return address (possibly fake or just unrelated) and the USPS tracking information (however much that's worth)), and an email exchange (sent from a nondescript gmail account on their end). Obviously the check can be validated to be fake, but will there be any way to pin it on a specific person?
Mondrian did some good representational art before he started with rectangles.
It sounds like the essential thing is freedom to change direction combined with a sort of good taste-- feeling a drive to explore and an ability to see what's promising.
Traditional Asian artists presumably followed a different trajectory, where they imitated for decades, but gradually developed their own style.
I wonder whether anyone started exploring after their hot streak.
This might be in contrast to Root-Bernstein's theory that scientists make important discoveries when they work on a practical problem which is adjacent to their field, with Pasteur as an example.
In thinking about the horrible performance of the FDA regarding COVID I was wondering if there is a country that has a good equivalent.
In thinking about Scott's anecdote about the kids unnecessarily dying because because of the FDA, when a viable alternative was available I was wondering why someone did not set up a clinic in a nearby country to give the treatment. Athletes used to go to Mexico to buy steroids all the time. Why can't people suffering from illnesses that have effective but illegal treatments do the same thing?
"Consider an isolated group of n students. They wish to assign themselves
unique student numbers from 1 to n (no repetitions!) in such a way that each
student will know his own student number, but each student should have no
information about any other student’s number. Devise a procedure for the
students to achieve their goal with 100% probability.
Some clarifications:
1 “isolated” means the students have no external aids of any kind (e.g.
no pencils or paper or hats); the only actions available to them are
talking to each other publicly or privately in groups of their choosing.
2 The students cannot conceal their real identities in any way. So if X
talks to Y , then Y recognizes X and Y recognizes X (for example, by
tone of voice, scent, etc.).
3 “no information” even excludes probabilistic information: student X
should have no estimates whatsoever about student Y ’s number except
that it is a number from 1 to n that is not his own.
4 You can assume that the students all want to achieve this goal, and
that they have “idealized minds” like in most thought experiments."
My extension is: why would this puzzle set a cryptographer's teeth on edge? What would they want the puzzle to demand instead? How would you solve that version, and how would you find a solution to this that's a long way from solving it?
What's a good way to buy a fairly but not exceedingly fancy engagement ring?
My girlfriend went ring shopping and found a ring she really likes for about $13k, which is maybe 25% more than I'd like to spend. How possible is it to bargain down a speciality store for a riing? I've also found comparable-ish rings - same specs on the diamond to within ~0.1 carat, same setting, same band - on bluenile.com for ~9K, but of course that's a website and I'm hesitant to buy something without seeing it in person.
I live in NYC, so if anyone has experience navigating the Diamond District, that would be very helpful!
(Not interested in "don't buy a ring!" arguments, but I could MAYBE be persuaded to look into lab diamonds)
My city floods frequently these days, and I sometimes find myself in a situation where I need to figure out a route across town which avoids flooded streets. As far as can tell, Google Maps does not help. The local news only helps in so far as telling me some major areas of town to avoid, but not specifically with individual streets. For instance, if I want to get to the airport, there may be a route there, but it often involves a lot of trial and error: heading down a road and then discovering it's underwater at some point, turning around, trying a different path, etc.
I basically rely on personal experience and memory to chooses my route, which is probably not the optimal solution.
Is there a way to use technology to solve this problem? Obviously, a crowd-sourcing solution could work, but does such a crowd sourcing solution currently exist?
Is there a clever way to infer which streets are flooded using Google Maps?
I tried quite a few antidepressants (and other pharmaceutical mood treatments) in the past before giving up on them. Recently, things have gotten to the point where I'm reconsidering getting treatment, particularly after coming across Dr. Ken Gillman's content online about how underrated MAOIs have become (to illustrate, apparently many people who fail ECT respond very well to MAOI antidepressants).
Needless to say, I'm interested in giving one of these meds a try. However, none of them are on the market where I live (a country in E. Europe), and I'm putting this out there hoping someone might know how I could get a prescription for one (and the medication itself), ideally without having to travel to another country.
I should mention I've already had a trial of Moclobemide (prescribed abroad, way before Covid), but AFAIK, that's considered a very weak antidepressant, likely owing to it not being an irreversible MAOI, like Nardil or Parnate.
Thank you if you read this and any feedback would be highly appreciated.
> Methane has 28x the warming potential of CO2 over a 100-year time horizon (https://cdiac.ess-dive.lbl.gov/pns/current_ghg.html) but I think it's like 80x over a 30-year horizon or something like that. A noteworthy result of this is that if our CH4 output is stable, then CH4 levels won't increase in the long run. By contrast, CO2 accumulates, so in the long run the CO2:CH4 ratio should continue increasing. If we look at the 100-year time horizon, then, natural gas producers would have to leak about 3.7% of all their CH4 in order to produce the same amount of global warming via leaking as they produce via burning (remember, coal produces about 2x the CO2, so a 3.7% leak would put it on par with coal in terms of greenhouse emissions).
Not quite following here. If you produce 0.5 times as much CH4 but it's 80 times as harmful, it seems like you would have to leak 1/(0.5*80)=1/40 = 2.5% of your CH4 to have the same warming. How do you arrive at 3.7%?
Also, any figures on how much is leaked in practice?
A Dutch court ruled that Uber is an employer, rather than a mediator, because they have a lot of control over how the drivers do their work. The consequence is that they have to provide fixed contracts, pay for pensions, etc, although another 'platform company' lost a similar case and didn't change anything; with new cases being in the courts, of people demanding a fixed contract.
I posted this in 188 just before thread 189 appeared; I hope it's okay if I repost for more exposure.
When governments spend more than they collect in taxes, they do something that everyone refers to as "borrowing", which increases the "debt". But during the Covid pandemic, pretty much every country was "borrowing". But if everyone is borrowing, who is the lender? It seems to me now that those words do not have their ordinary meaning. "Borrowing" turns out to be code for "printing money", and "debt" is "the amount of money we've printed".
Well, not quite. I have the impression that governments nominally "borrow" from private corporations and individuals, but of course the way they "pay back" this money - with interest - is not by intermittently switching between budget deficit and budget surplus. Rather, they simply "borrow" more even money and use the new money to pay off the old debts. Which makes no sense to me: wouldn't it be better to print money to avoid paying interest? (and to avoid the risk of hyperinflation, have some sort of limit on the money-printing?)
I wonder if the whole system is set up in some modestly idiotic way - inefficient and difficult to understand, but not bad enough that the government is forced to change it. I also wonder if all the governments of the world use basically the same system, which would be a surprising "coincidence".
In any case, I've never seen a explanation that I could entirely follow. Aside from things like "fractional reserve" being hard to wrap one's head around, I find that virtually everyone who tries to explain macroeconomics takes for granted that their audience understands concepts like "buying debt" and the distinction between "fiscal", "monetary" and "financial". Does anyone explain this stuff like I'm 5?
Still, I would like to share a flash of insight I've had recently about macroeconomics that no one has ever even attempted to explain to me. It's about the value of money.
I assert that the value of money is (approximately) the total amount of production divided by the total amount of spending. "Amount of production" is real-world goods and services, so it has no particular unit of measurement. "Amount of spending" is the amount of money that changes hands, and it could be measured in dollars.
A key point here is that money which doesn't change hands doesn't enter into the equation. Nor does the world population. Hypothetically, then, suppose Jeff Bezos finds a way to gobble up most of the world's wealth and he becomes a 50-trillionaire. If production stays the same during this time (I guess it's more likely to increase, but let's pretend) and he spends almost none of this money, the effect of this wealth accumulation should be deflationary: the denominator (spending) decreases because Bezos is not spending his earnings (while production is flat or increasing), so the value of money increases. Everyone's money is worth more! Yay! However, those who are in debt effectively find themselves with bigger debts. Wages fall in response to the constricted money supply, so indebted people will have trouble paying off their debts. (I heard somewhere that this was a major problem during the Great Depression.)
But now, suppose that suddenly Bezos decides to spend 6 trillion dollars for a vacation on the moon three years from now, and suppose world production responds mainly by *moving* resources to the moon mission (due to structural limitations that prevent total production from increasing very much). Thanks to the increasing denominator, the effect will be sudden inflation (especially in moon-mission-related industries, i.e. this is where price increases are likely to be concentrated, though there will be inflation everywhere due to the loss of production in other sectors, and also whole supply chains relevant to the moon mission will be impacted, which can also cause price increases to bleed into other areas of the economy).
Getting back to the debt issue, while nominally the U.S. and many other countries have huge public debts and also huge private debts, none of this matters in practice, *as long as it doesn't affect production or spending significantly*. Indeed, perhaps big debts can be good by stimulating production, though I wonder if it can lead to instability (and if so, why).
Also, anyone want to predict the overall stock market trend over the next few years? I am not aware of any mechanism by which a major crash should occur, so I tentatively expect a minor crash at worst. However, US stocks are probably overpriced, so I expect that price increases will level off pretty soon and investor returns will be relatively poor over the next few years. Of course, though, I'm no expert and I don't really understand why the stock market rose so much in the first place. Did a lot of those stimulus dollars somehow get dumped straight into markets? Or was it caused more by regulators using their poorly-explained mechanisms to increase the money supply in a way that increased average stock prices?
Have recently mocked a work of fiction for having a secret organisation (i.e., no-one knows it exists) with 100,000,000 members. I feel fairly confident in this judgement, but I'm no expert in such matters and it has occurred to me that I don't really have a good feel for how large a conspiracy could actually plausibly be.
What would be personal estimates for the largest plausible size of a fully-secret organisation (nobody knows it exists; can't think of easy past examples), and of the largest plausible size of a publically-known organisation's secret projects (nobody knows that organisation does that kind of thing; Dragonfly would be an example of this before it was leaked)?
Hello all, I am a graduate student in geometry (Riemannian, Conformal, CR), who is looking for a mathematical collaborator. I feel that I have a willingness to learn new things, and am committed to being a helpful and pleasant collaborator. My desire for a collaborator stems from this recent article on Slime Mold Time Mold- https://www.strangeloopcanon.com/p/why-we-fund-teams-and-hire-individuals
If you are someone who is interested in collaborating with a grad student interested in geometry, you may reply here or mail me at auk480@psu.edu. For context, my research will be uploaded on arXiv soon (when I am done removing them blasted typos from my preprints).
I was reading about SETI (Search for Extraterrestrial Intelligence) and a question popped into my head - hopefully the experts here can weigh in. If SETI were happening on one of the exoplanets closest to us (in the Proxima Centauri system 4+ light years away), would they be able to detect the signals from Earth? If yes - could they, with the appropriate technology, listen to the top 40 radio hits of 4 years ago? If no - what is order of magnitude by which it fails? 2x, 10x, 100x? And then why are people doing something that can't work?
If Larry Elder wins the second Q vote by a large margin as seems likely, will you count the "Why aren't you governor" post as a mistake?
Incidentally, Newsom's strategy of ensuring that no serious Democrats run so it would be a partisan matchup seems like it's working great to me. I don't know why everyone kept criticizing it.
i had a routine colonoscopy done a few months ago as a cancer screen and it came back with no cancer but some inflammation. i had a few tests (stool and blood) which showed slightly elevated inflammation markers. my gastroenterologist diagnosed me with colitis, although i am not aware that i have any symptoms. i think my digestion is fine. he told me that i should take mesalamine daily for the rest of my life to forestall a flare up. i haven't been able to find evidence one way or the other that this is necessary. does anyone know whether this is necessary?
At college I studied mathematics and philosophy. I don't entirely regret studying philosophy, and I love talking to my philosopher friends; in terms of personality and interests, I have much more in common with them than with the people I made friends with when I went into bioscience.
But nowadays it is absolutely the case that I try to answer the philosophical questions I have with reference to scientific findings. I wouldn't say you can solve all philosophical questions from knowing these things. But I would say that answering questions of interest to philosophers needs to be built on the science we know, and it's troubling to me that there are entire fields of philosophy that don't seem to care about pertinent findings.
I've made a few lists of "scientific findings I'd teach philosophy undergrads", here's one:
-Primatology and evolutionary psychology
-Evolutionary game theory, especially costly signalling
-Genetics and its relation to neuroscience
-Bayesian updating
-Information theory
-Godel's Incompleteness Theorems and Turing-completeness
-Basic special relativity and quantum mechanics
It's an old debate I know, and this list maybe says more about me (and lesswrong) than about the world. But I'd be interested to hear how others feel about where we are with this.
We've all heard the anecdotes about production in Soviet factories going awry due to poorly devised incentive schemes. In a chandelier factory, for example, where workers were paid a bonus based on the tonnage of output, chandeliers grew heavier and heavier until they were too heavy to remain attached to the ceilings.
A similar story involves production quotas for nail factories. To meet the quota at the lowest possible costs, the story goes, factories produced thousands of super small and useless nails.
It occurs to me that these stories (if true) are examples of intelligent systems, where the managers involved presumably understand the ultimate goal of the incentive designer (efficiently produce useful chandeliers and nails), and nonetheless opt to cheat the incentive system and ignore the goals of the designer. In the same way that an AGI might understand the ultimate goals of its designer but choose to ignore them because they are not aligned with its utility function.
Anyway, I thought this was an interesting idea. These examples can be used as a response to the argument that a smart enough AGI would not choose actions that go against its designer's interests, as it could use common sense to infer what the designer's ultimate goals. Even humans, who definitely understand the goals that the incentive designer wants to achieve, cannot be trusted to ignore their utility function in pursue of said goal.
Practical issues with approval voting for general elections: (I'm assuming this is OK in the 'politics free' open thread as it's a fairly dry discussion of electoral methods, not a culture war post- if it's not OK I'll delete). This is assuming that you already have a working knowledge of approval voting https://en.wikipedia.org/wiki/Approval_voting
1. In practice, it seems that the large majority of voters 'bullet vote', or simply cast one vote/approval (as opposed to 'approving' multiple candidates). While this is fine and doesn't break the system, it does undermine the supposed advantage of AV- that you're ultimately getting a consensus candidate because voters can approve multiple people. In St. Louis' recent mayoral election, voters averaged less than 1.5 approvals per ballot- the Mathematical Association of America stopped using AV because around 80% of voters would only cast 1 ballot. The IEEE (a pretty prominent organization!) ceased using AV after several years because reportedly 'few of our members were using it'. At Dartmouth University their alumni association stopped using AV as they also cited 80% of voters were bullet voting.
Again, nothing wrong with bullet voting- but you're simply doing FPTP at that point.
2. AV is arguably inferior to simple two round FPTP at handling multiple candidates. With enough candidates running, the winner can have a plurality that's still a small minority of all voters (of course this combines with issue 1 in that if most voters are bullet voting, anything above 4 or 5 people running will result in just a small plurality). Again at Dartmouth University, out of a large field for the presidency of their Alumni Association, the winner won with 32% of the vote in 2012, and under 40% in 3 other years.
With FPTP plus a runoff/two round system, this by definition isn't an issue. This is a not pro-ranked choice voting comment BTW- I have even more issues with RCV!- and I think AV has some real potential for the specific use case of party primaries. For general elections though, a two round runoff is probably the most practical system. There's a reason 49 democracies worldwide use it, and none of them use approval voting, despite the fact that AV has been around for hundreds of years!
"Effective altruist organization Open Phil is offering scholarships to interested-in-effective-altruism international undergrads applying to top US and UK universities."
How the hell do you call yourself an effective altruist organization when you take money and give it to top US and UK universities instead of investing it in universities that do much more to change their students lives? Elite Universities are already over-endowed, while the schools doing the most to improve their students financial status struggle with funding.
Not the mention the brain-drain element. Instead of sending Colombians to the US for education, why not work to make Colombian universities better? It's not like there is a shortage of PhDs who could do amazing work in a university in Kenya.
If you're tired of being mad at the FDA, why not take a break and get mad at the FRA instead?
There is a private company (Texas Central) that is trying to build high speed rail between Dallas and Houston. They are using cheaper Japanese trains, even though that means that they can't use any existing railroad tracks. The Federal Railroad Administration prefers European trains and is mad at this:
"Of note, the FRA speaks of grade crossings on a line that has none, and demands trains to withstand the impact of a 6.35 ton steel ball that may be dropped from overpasses that do not exist. This is likely malicious more than incompetent ..."
From the discussion from the last thread, the most obvious candidates for the fastest declining major languages of the twentieth century have been the classical languages -Sanskrit, Latin, and literary Chinese. Does anybody disagree?
Food Insecurity: Does anyone know of any studies that look at two similar groups afflicted by food insecurity, make one group food secure and then follow both groups over time to see what the payoff is?
There are ethical issues inherent in my question, but I'm actively engaged in helping to to provide food to a group of families that are food insecure, and I am wondering if there is any way to test the cost.benefit against some sort of baseline.
The recent post on semiheavy water reminded me of a concern I've had for a while, but never got around to articulating. It's a bit hard to explain, and I'm not the greatest at explaining things (at least I don't think I am), but I'll try to do the best I can.
In Scott's famous post "The Control Group Is Out Of Control", Scott presents the idea of what we might call "meta-control-grouping". In a drug trial, you have a control group that you give a drug you *know* doesn't work, and this shows you how many people will get an effect even if the drug doesn't work; and unless the drug does better than control, you have no evidence the drug does anything. Similarly, on a meta level, if there were a scientific effect that you *knew* didn't exist, you could use that as a meta-control-group, and see what kinds of results you would get even if the effect didn't exist. Unless your experiment does better than the meta-control, you have no evidence the effect exists.
Scott uses the example of parapsychology as a meta control group; and while I'm far less certain than he is that parapsychological effects don't exist (on some days I even lean in favor of them existing - needless to say my priors are much different than his), the overall idea seems intuitively sound. If an individual or community has very low priors on phenomenon A, but phenomenon A is getting evidence E(A) as good as or better than the evidence E(B) for a phenomenon B that the individual or community has high priors on, this should call into question the validity of E(B).
But this is hard for me to square with what I know about Bayesianism in general. The whole point of Bayesianism is that "extraordinary claims require extraordinary evidence": E(B) is good enough for B while the same level of evidence E(A) is not good enough for A, precisely because you have such low priors for A. For example, if both E(A) and E(B) give you 3 bits of evidence, then 3 bits might be enough to push you above 50% for B (because you were already close to 50%), but not enough to push you above 50% for A (because you were already so far below 50%).
But the idea of meta control grouping seems to rule out "double standards" like this. What the idea of meta control grouping seems to imply is that, because you got 3 bits of evidence for A, and A is likely false, that means you can easily get 3 bits of evidence for a false proposition, and thus 3 bits of evidence is actually not evidence at all (or at least much less evidence than you thought). So what this would mean is, from now on, any time you get 3 bits of evidence for anything, you should either discount it, or "correct" it down to (say) 1 bit of evidence. For the sake of communication, I will call this idea Devaluation of Evidence (DE).
I can see this getting out of control very quickly. Any time you get evidence for a proposition you have low priors on, that amount of evidence becomes worthless for every other proposition you've ever believed. This should make you progressively more uncertain about everything over time, until (in the limit) every one of your beliefs is 50-50 and you have no beliefs about anything whatsoever. For the sake of ease of communication, I will call this idea the Progressive Devaluation of Evidence (PDE).
In fact, I almost want to say that this results in an infinite loop right off the bat. If 3 bits in favor of B needs to be corrected down to 1 bit in favor of B, then does that same 1 bit in favor of B need to be corrected down to 1/3 of a bit in favor of B, and then down to 1/9 of a bit in favor of B, and so on? I'm far less sure about this, but this seems to be a potential issue as well. (Let's call this the Devaluation of Evidence Infinite Loop [DEIL]).
In other words, unless I'm misunderstanding something, the idea of meta-control-grouping seems to break Bayesianism. I'm not certain about any of this, and I get the feeling I'm fundamentally misunderstanding something here, but I just wanted to get this out so that someone knowledgable can clear this up for me.
Something I wish people would focus more attention on, when discussing "privilege" and "oppression", is the effect of habituation and the hedonic treadmill. After all, just because there's a relative difference between two people it doesn't automatically imply a difference in perceived satisfaction. People very quickly get habituated with their life situation, and as suicidal celebrities discover, having stuff doesn't automatically imply you're leading a happier life.
I'm curious to what extent various social and psychological ills would turn out, upon closer and honest inspection, to be self-fulfilling prophecies. That is, the fact that people are made aware they should "feel bad" is actually the main cause of them experiencing negative emotions in the first place.
For one thing, I'm thinking about various "traumas" that people tend to get groomed into believing they've experienced, often in a cult-like manner. But I also have in mind the psychological effects of "inequality" (which are often the effects of becoming aware of, constantly perceiving, and obsessing over inequality, more than of the actual inequality itself).
The approved universities for the EA scholarship seem to be particularly elitism and classist for a community so aligned with Rationalist thought. I'd understand some combination of planned course of study, explanation, need and minimally prestigious university (which goes towards ability to change the world for the better). Instead it is limited to a particularly tiny set of schools. Maybe I'm making a molehill out of an anthill here, but this is the same flavoring that has put me off EA forums in the past.
I recommend reading the whole very short piece, the choice bits:
"We are going to punish somebody for this attack, but just who or what will be blown to smithereens for it is hard to say. Maybe Afghanistan, maybe Pakistan or Iraq, or possibly all three at once. Who knows? Not even the Generals in what remains of the Pentagon or the New York papers calling for WAR seem to know who did it or where to look for them."
"The towers are gone now, reduced to bloody rubble, along with all hopes for Peace in Our Time, in the United States or any other country. Make no mistake about it: We are At War now -- with somebody -- and we will stay At War with that mysterious Enemy for the rest of our lives."
To recap, this is when you come up with a rational sounding reason for some opinion/action you want to defend.
I’ve long been aware of it, but I didn’t understand it is typically an unconscious process: Some optimizing part of your mind decides you want X. The Motivated Reasoning part then comes up with as good a sounding argument for why this is the just and right thing to happen. And then you - the tip of the iceberg conscious “you” - ACTUALLY BELIEVE IT!
This has some implications
Often people will argue using obviously broken or dishonest arguments. I used to think they were lying. But now I think they’re maybe BEING LIED TO, by their Motivated Reasoning Center.
This must also imply different methods of arguing with such people. The “come on, admit you’re lying” strategy is obviously out. What, if anything, are some better ways?
Hey all! I forget if this violates any rules about self-promotion on open threads (if so just let me know and I’ll delete) but I’ve recently been investigating how AI-generated art and text is draining away meaning. I think this is a big problem that people, especially the creative class, will be blindsided by. What happens when all music is generated by some non-conscious Chinese Room? I have no idea but I find it terrifying to consider. Any thoughts (or reassurances) are welcome https://erikhoel.substack.com/p/big-tech-is-replacing-human-artists
Magic 8 Ball like bullet point from FDA Panel meeting today:
“The timing for possible Moderna and Johnson & Johnson booster in U.S. is murky.”
In the frustrating vagueness department that's up there with the actually Magic 8 Ball response: "Reply Hazy. Try again later."
Jebus.
Scott should do an epilogue to the "Too Good to Check" post, starring Nicki Minaj
according to this
https://files.givewell.org/files/ClearFund/Meeting_2020_08_25/Attachment_F_DRAFT_GiveWell_Financial_Summary___August_25_2020_2.pdf
givewell spend about 1/3 of its revenue on expense before sending it to charity
mostly on staff
but 1.2M USD is spent on admin from 18M revenue
this does not seem effective.
is it true or did I misread it?
I've had a scammer try to lead me through a cashiers check scam. (Briefly: I was selling an item on Craigslist, they said they'd mail me a check and could I pay the movers out of it. The scam continues with "the check is fake but it's a cashier's check so the bank gives you the money now, then someone -- likely you -- has to pony up when it's revealed the check is fake after all".) Not knowing the scam beforehand, I played along one step farther than I would've liked, and am currently in possession of the (presumed fake) check.
Question 1: is there some way to punish this behavior? I haven't been materially harmed (yet; the scammer does know an uncomfortable amount of information about me), but presumably writing fake checks is illegal even if they're not cashed. But, who would I report this to? My police department? "Their" police department (based on the address)?
Question 2: is there even enough information for anyone to do anything about it? I have a cashier's check (allegedly from a small business, not a person), a letter by which it arrived (which has a return address (possibly fake or just unrelated) and the USPS tracking information (however much that's worth)), and an email exchange (sent from a nondescript gmail account on their end). Obviously the check can be validated to be fake, but will there be any way to pin it on a specific person?
https://phys.org/news/2021-09-secret-van-gogh-success.html
Of course, it isn't about the secret of van Gogh's success It's about a deep data dive into careers in art, film direction, and science.
They found a period of exploration followed by exploitation when a promising field was found.
https://www.nature.com/articles/s41467-021-25477-8
Mondrian did some good representational art before he started with rectangles.
It sounds like the essential thing is freedom to change direction combined with a sort of good taste-- feeling a drive to explore and an ability to see what's promising.
Traditional Asian artists presumably followed a different trajectory, where they imitated for decades, but gradually developed their own style.
I wonder whether anyone started exploring after their hot streak.
This might be in contrast to Root-Bernstein's theory that scientists make important discoveries when they work on a practical problem which is adjacent to their field, with Pasteur as an example.
In thinking about the horrible performance of the FDA regarding COVID I was wondering if there is a country that has a good equivalent.
In thinking about Scott's anecdote about the kids unnecessarily dying because because of the FDA, when a viable alternative was available I was wondering why someone did not set up a clinic in a nearby country to give the treatment. Athletes used to go to Mexico to buy steroids all the time. Why can't people suffering from illnesses that have effective but illegal treatments do the same thing?
Here's a fun puzzle I found on the internet, by a guy called Andrew Critch, at http://www.acritch.com/media/math/Self-assigned_student_numbers.pdf.
His puzzle is:
"Consider an isolated group of n students. They wish to assign themselves
unique student numbers from 1 to n (no repetitions!) in such a way that each
student will know his own student number, but each student should have no
information about any other student’s number. Devise a procedure for the
students to achieve their goal with 100% probability.
Some clarifications:
1 “isolated” means the students have no external aids of any kind (e.g.
no pencils or paper or hats); the only actions available to them are
talking to each other publicly or privately in groups of their choosing.
2 The students cannot conceal their real identities in any way. So if X
talks to Y , then Y recognizes X and Y recognizes X (for example, by
tone of voice, scent, etc.).
3 “no information” even excludes probabilistic information: student X
should have no estimates whatsoever about student Y ’s number except
that it is a number from 1 to n that is not his own.
4 You can assume that the students all want to achieve this goal, and
that they have “idealized minds” like in most thought experiments."
My extension is: why would this puzzle set a cryptographer's teeth on edge? What would they want the puzzle to demand instead? How would you solve that version, and how would you find a solution to this that's a long way from solving it?
What's a good way to buy a fairly but not exceedingly fancy engagement ring?
My girlfriend went ring shopping and found a ring she really likes for about $13k, which is maybe 25% more than I'd like to spend. How possible is it to bargain down a speciality store for a riing? I've also found comparable-ish rings - same specs on the diamond to within ~0.1 carat, same setting, same band - on bluenile.com for ~9K, but of course that's a website and I'm hesitant to buy something without seeing it in person.
I live in NYC, so if anyone has experience navigating the Diamond District, that would be very helpful!
(Not interested in "don't buy a ring!" arguments, but I could MAYBE be persuaded to look into lab diamonds)
How do I figure out which streets are flooded?
My city floods frequently these days, and I sometimes find myself in a situation where I need to figure out a route across town which avoids flooded streets. As far as can tell, Google Maps does not help. The local news only helps in so far as telling me some major areas of town to avoid, but not specifically with individual streets. For instance, if I want to get to the airport, there may be a route there, but it often involves a lot of trial and error: heading down a road and then discovering it's underwater at some point, turning around, trying a different path, etc.
I basically rely on personal experience and memory to chooses my route, which is probably not the optimal solution.
Is there a way to use technology to solve this problem? Obviously, a crowd-sourcing solution could work, but does such a crowd sourcing solution currently exist?
Is there a clever way to infer which streets are flooded using Google Maps?
Update your cryonics and/or Russia related priors accordingly: https://www.thetimes.co.uk/article/cryogenics-tycoon-danila-medvedev-accuses-ex-wife-of-stealing-frozen-bodies-g6mmbl3h3 (please excuse the source, but it's the best coverage I saw)
Considering the ivermectin/gunshot wounds mess, how much trust should I put in this?
https://www.npr.org/sections/coronavirus-live-updates/2021/09/13/1036593269/coronavirus-alabama-43-icus-at-capacity-ray-demonia
I tried quite a few antidepressants (and other pharmaceutical mood treatments) in the past before giving up on them. Recently, things have gotten to the point where I'm reconsidering getting treatment, particularly after coming across Dr. Ken Gillman's content online about how underrated MAOIs have become (to illustrate, apparently many people who fail ECT respond very well to MAOI antidepressants).
Needless to say, I'm interested in giving one of these meds a try. However, none of them are on the market where I live (a country in E. Europe), and I'm putting this out there hoping someone might know how I could get a prescription for one (and the medication itself), ideally without having to travel to another country.
I should mention I've already had a trial of Moclobemide (prescribed abroad, way before Covid), but AFAIK, that's considered a very weak antidepressant, likely owing to it not being an irreversible MAOI, like Nardil or Parnate.
Thank you if you read this and any feedback would be highly appreciated.
Continuing the [fracking vs. Coal discussion](https://astralcodexten.substack.com/p/open-thread-188/comments#comment-2821624)
> Methane has 28x the warming potential of CO2 over a 100-year time horizon (https://cdiac.ess-dive.lbl.gov/pns/current_ghg.html) but I think it's like 80x over a 30-year horizon or something like that. A noteworthy result of this is that if our CH4 output is stable, then CH4 levels won't increase in the long run. By contrast, CO2 accumulates, so in the long run the CO2:CH4 ratio should continue increasing. If we look at the 100-year time horizon, then, natural gas producers would have to leak about 3.7% of all their CH4 in order to produce the same amount of global warming via leaking as they produce via burning (remember, coal produces about 2x the CO2, so a 3.7% leak would put it on par with coal in terms of greenhouse emissions).
Not quite following here. If you produce 0.5 times as much CH4 but it's 80 times as harmful, it seems like you would have to leak 1/(0.5*80)=1/40 = 2.5% of your CH4 to have the same warming. How do you arrive at 3.7%?
Also, any figures on how much is leaked in practice?
A Dutch court ruled that Uber is an employer, rather than a mediator, because they have a lot of control over how the drivers do their work. The consequence is that they have to provide fixed contracts, pay for pensions, etc, although another 'platform company' lost a similar case and didn't change anything; with new cases being in the courts, of people demanding a fixed contract.
It is still possible for Uber to appeal.
U.S. parents: how did you choose the preschool / kindergarten / elementary school your child attended? What value(s) informed your decision?
I posted this in 188 just before thread 189 appeared; I hope it's okay if I repost for more exposure.
When governments spend more than they collect in taxes, they do something that everyone refers to as "borrowing", which increases the "debt". But during the Covid pandemic, pretty much every country was "borrowing". But if everyone is borrowing, who is the lender? It seems to me now that those words do not have their ordinary meaning. "Borrowing" turns out to be code for "printing money", and "debt" is "the amount of money we've printed".
Well, not quite. I have the impression that governments nominally "borrow" from private corporations and individuals, but of course the way they "pay back" this money - with interest - is not by intermittently switching between budget deficit and budget surplus. Rather, they simply "borrow" more even money and use the new money to pay off the old debts. Which makes no sense to me: wouldn't it be better to print money to avoid paying interest? (and to avoid the risk of hyperinflation, have some sort of limit on the money-printing?)
I wonder if the whole system is set up in some modestly idiotic way - inefficient and difficult to understand, but not bad enough that the government is forced to change it. I also wonder if all the governments of the world use basically the same system, which would be a surprising "coincidence".
In any case, I've never seen a explanation that I could entirely follow. Aside from things like "fractional reserve" being hard to wrap one's head around, I find that virtually everyone who tries to explain macroeconomics takes for granted that their audience understands concepts like "buying debt" and the distinction between "fiscal", "monetary" and "financial". Does anyone explain this stuff like I'm 5?
Still, I would like to share a flash of insight I've had recently about macroeconomics that no one has ever even attempted to explain to me. It's about the value of money.
I assert that the value of money is (approximately) the total amount of production divided by the total amount of spending. "Amount of production" is real-world goods and services, so it has no particular unit of measurement. "Amount of spending" is the amount of money that changes hands, and it could be measured in dollars.
A key point here is that money which doesn't change hands doesn't enter into the equation. Nor does the world population. Hypothetically, then, suppose Jeff Bezos finds a way to gobble up most of the world's wealth and he becomes a 50-trillionaire. If production stays the same during this time (I guess it's more likely to increase, but let's pretend) and he spends almost none of this money, the effect of this wealth accumulation should be deflationary: the denominator (spending) decreases because Bezos is not spending his earnings (while production is flat or increasing), so the value of money increases. Everyone's money is worth more! Yay! However, those who are in debt effectively find themselves with bigger debts. Wages fall in response to the constricted money supply, so indebted people will have trouble paying off their debts. (I heard somewhere that this was a major problem during the Great Depression.)
But now, suppose that suddenly Bezos decides to spend 6 trillion dollars for a vacation on the moon three years from now, and suppose world production responds mainly by *moving* resources to the moon mission (due to structural limitations that prevent total production from increasing very much). Thanks to the increasing denominator, the effect will be sudden inflation (especially in moon-mission-related industries, i.e. this is where price increases are likely to be concentrated, though there will be inflation everywhere due to the loss of production in other sectors, and also whole supply chains relevant to the moon mission will be impacted, which can also cause price increases to bleed into other areas of the economy).
Getting back to the debt issue, while nominally the U.S. and many other countries have huge public debts and also huge private debts, none of this matters in practice, *as long as it doesn't affect production or spending significantly*. Indeed, perhaps big debts can be good by stimulating production, though I wonder if it can lead to instability (and if so, why).
Also, anyone want to predict the overall stock market trend over the next few years? I am not aware of any mechanism by which a major crash should occur, so I tentatively expect a minor crash at worst. However, US stocks are probably overpriced, so I expect that price increases will level off pretty soon and investor returns will be relatively poor over the next few years. Of course, though, I'm no expert and I don't really understand why the stock market rose so much in the first place. Did a lot of those stimulus dollars somehow get dumped straight into markets? Or was it caused more by regulators using their poorly-explained mechanisms to increase the money supply in a way that increased average stock prices?
Have EAers responded to the critiques in this essay by Crary? https://www.radicalphilosophy.com/article/against-effective-altruism
I'm assuming yes, but it would be neat to know whether there is a one-stop-shop of responses.
Reference: https://slatestarcodex.com/2019/01/14/too-many-people-dare-call-it-conspiracy/
Have recently mocked a work of fiction for having a secret organisation (i.e., no-one knows it exists) with 100,000,000 members. I feel fairly confident in this judgement, but I'm no expert in such matters and it has occurred to me that I don't really have a good feel for how large a conspiracy could actually plausibly be.
What would be personal estimates for the largest plausible size of a fully-secret organisation (nobody knows it exists; can't think of easy past examples), and of the largest plausible size of a publically-known organisation's secret projects (nobody knows that organisation does that kind of thing; Dragonfly would be an example of this before it was leaked)?
Advertisement: For a mathematical collaborator.
Hello all, I am a graduate student in geometry (Riemannian, Conformal, CR), who is looking for a mathematical collaborator. I feel that I have a willingness to learn new things, and am committed to being a helpful and pleasant collaborator. My desire for a collaborator stems from this recent article on Slime Mold Time Mold- https://www.strangeloopcanon.com/p/why-we-fund-teams-and-hire-individuals
If you are someone who is interested in collaborating with a grad student interested in geometry, you may reply here or mail me at auk480@psu.edu. For context, my research will be uploaded on arXiv soon (when I am done removing them blasted typos from my preprints).
I was reading about SETI (Search for Extraterrestrial Intelligence) and a question popped into my head - hopefully the experts here can weigh in. If SETI were happening on one of the exoplanets closest to us (in the Proxima Centauri system 4+ light years away), would they be able to detect the signals from Earth? If yes - could they, with the appropriate technology, listen to the top 40 radio hits of 4 years ago? If no - what is order of magnitude by which it fails? 2x, 10x, 100x? And then why are people doing something that can't work?
If Larry Elder wins the second Q vote by a large margin as seems likely, will you count the "Why aren't you governor" post as a mistake?
Incidentally, Newsom's strategy of ensuring that no serious Democrats run so it would be a partisan matchup seems like it's working great to me. I don't know why everyone kept criticizing it.
i had a routine colonoscopy done a few months ago as a cancer screen and it came back with no cancer but some inflammation. i had a few tests (stool and blood) which showed slightly elevated inflammation markers. my gastroenterologist diagnosed me with colitis, although i am not aware that i have any symptoms. i think my digestion is fine. he told me that i should take mesalamine daily for the rest of my life to forestall a flare up. i haven't been able to find evidence one way or the other that this is necessary. does anyone know whether this is necessary?
At college I studied mathematics and philosophy. I don't entirely regret studying philosophy, and I love talking to my philosopher friends; in terms of personality and interests, I have much more in common with them than with the people I made friends with when I went into bioscience.
But nowadays it is absolutely the case that I try to answer the philosophical questions I have with reference to scientific findings. I wouldn't say you can solve all philosophical questions from knowing these things. But I would say that answering questions of interest to philosophers needs to be built on the science we know, and it's troubling to me that there are entire fields of philosophy that don't seem to care about pertinent findings.
I've made a few lists of "scientific findings I'd teach philosophy undergrads", here's one:
-Primatology and evolutionary psychology
-Evolutionary game theory, especially costly signalling
-Genetics and its relation to neuroscience
-Bayesian updating
-Information theory
-Godel's Incompleteness Theorems and Turing-completeness
-Basic special relativity and quantum mechanics
It's an old debate I know, and this list maybe says more about me (and lesswrong) than about the world. But I'd be interested to hear how others feel about where we are with this.
We've all heard the anecdotes about production in Soviet factories going awry due to poorly devised incentive schemes. In a chandelier factory, for example, where workers were paid a bonus based on the tonnage of output, chandeliers grew heavier and heavier until they were too heavy to remain attached to the ceilings.
A similar story involves production quotas for nail factories. To meet the quota at the lowest possible costs, the story goes, factories produced thousands of super small and useless nails.
It occurs to me that these stories (if true) are examples of intelligent systems, where the managers involved presumably understand the ultimate goal of the incentive designer (efficiently produce useful chandeliers and nails), and nonetheless opt to cheat the incentive system and ignore the goals of the designer. In the same way that an AGI might understand the ultimate goals of its designer but choose to ignore them because they are not aligned with its utility function.
Anyway, I thought this was an interesting idea. These examples can be used as a response to the argument that a smart enough AGI would not choose actions that go against its designer's interests, as it could use common sense to infer what the designer's ultimate goals. Even humans, who definitely understand the goals that the incentive designer wants to achieve, cannot be trusted to ignore their utility function in pursue of said goal.
Practical issues with approval voting for general elections: (I'm assuming this is OK in the 'politics free' open thread as it's a fairly dry discussion of electoral methods, not a culture war post- if it's not OK I'll delete). This is assuming that you already have a working knowledge of approval voting https://en.wikipedia.org/wiki/Approval_voting
1. In practice, it seems that the large majority of voters 'bullet vote', or simply cast one vote/approval (as opposed to 'approving' multiple candidates). While this is fine and doesn't break the system, it does undermine the supposed advantage of AV- that you're ultimately getting a consensus candidate because voters can approve multiple people. In St. Louis' recent mayoral election, voters averaged less than 1.5 approvals per ballot- the Mathematical Association of America stopped using AV because around 80% of voters would only cast 1 ballot. The IEEE (a pretty prominent organization!) ceased using AV after several years because reportedly 'few of our members were using it'. At Dartmouth University their alumni association stopped using AV as they also cited 80% of voters were bullet voting.
Again, nothing wrong with bullet voting- but you're simply doing FPTP at that point.
2. AV is arguably inferior to simple two round FPTP at handling multiple candidates. With enough candidates running, the winner can have a plurality that's still a small minority of all voters (of course this combines with issue 1 in that if most voters are bullet voting, anything above 4 or 5 people running will result in just a small plurality). Again at Dartmouth University, out of a large field for the presidency of their Alumni Association, the winner won with 32% of the vote in 2012, and under 40% in 3 other years.
With FPTP plus a runoff/two round system, this by definition isn't an issue. This is a not pro-ranked choice voting comment BTW- I have even more issues with RCV!- and I think AV has some real potential for the specific use case of party primaries. For general elections though, a two round runoff is probably the most practical system. There's a reason 49 democracies worldwide use it, and none of them use approval voting, despite the fact that AV has been around for hundreds of years!
"Effective altruist organization Open Phil is offering scholarships to interested-in-effective-altruism international undergrads applying to top US and UK universities."
How the hell do you call yourself an effective altruist organization when you take money and give it to top US and UK universities instead of investing it in universities that do much more to change their students lives? Elite Universities are already over-endowed, while the schools doing the most to improve their students financial status struggle with funding.
Not the mention the brain-drain element. Instead of sending Colombians to the US for education, why not work to make Colombian universities better? It's not like there is a shortage of PhDs who could do amazing work in a university in Kenya.
If you're tired of being mad at the FDA, why not take a break and get mad at the FRA instead?
There is a private company (Texas Central) that is trying to build high speed rail between Dallas and Houston. They are using cheaper Japanese trains, even though that means that they can't use any existing railroad tracks. The Federal Railroad Administration prefers European trains and is mad at this:
"Of note, the FRA speaks of grade crossings on a line that has none, and demands trains to withstand the impact of a 6.35 ton steel ball that may be dropped from overpasses that do not exist. This is likely malicious more than incompetent ..."
https://pedestrianobservations.com/2021/06/24/are-the-fra-and-european-operators-sabotaging-texas-central/
From the discussion from the last thread, the most obvious candidates for the fastest declining major languages of the twentieth century have been the classical languages -Sanskrit, Latin, and literary Chinese. Does anybody disagree?
Does anyone know why the FDA is preventing cheap covid tests?
Food Insecurity: Does anyone know of any studies that look at two similar groups afflicted by food insecurity, make one group food secure and then follow both groups over time to see what the payoff is?
There are ethical issues inherent in my question, but I'm actively engaged in helping to to provide food to a group of families that are food insecure, and I am wondering if there is any way to test the cost.benefit against some sort of baseline.
The recent post on semiheavy water reminded me of a concern I've had for a while, but never got around to articulating. It's a bit hard to explain, and I'm not the greatest at explaining things (at least I don't think I am), but I'll try to do the best I can.
In Scott's famous post "The Control Group Is Out Of Control", Scott presents the idea of what we might call "meta-control-grouping". In a drug trial, you have a control group that you give a drug you *know* doesn't work, and this shows you how many people will get an effect even if the drug doesn't work; and unless the drug does better than control, you have no evidence the drug does anything. Similarly, on a meta level, if there were a scientific effect that you *knew* didn't exist, you could use that as a meta-control-group, and see what kinds of results you would get even if the effect didn't exist. Unless your experiment does better than the meta-control, you have no evidence the effect exists.
Scott uses the example of parapsychology as a meta control group; and while I'm far less certain than he is that parapsychological effects don't exist (on some days I even lean in favor of them existing - needless to say my priors are much different than his), the overall idea seems intuitively sound. If an individual or community has very low priors on phenomenon A, but phenomenon A is getting evidence E(A) as good as or better than the evidence E(B) for a phenomenon B that the individual or community has high priors on, this should call into question the validity of E(B).
But this is hard for me to square with what I know about Bayesianism in general. The whole point of Bayesianism is that "extraordinary claims require extraordinary evidence": E(B) is good enough for B while the same level of evidence E(A) is not good enough for A, precisely because you have such low priors for A. For example, if both E(A) and E(B) give you 3 bits of evidence, then 3 bits might be enough to push you above 50% for B (because you were already close to 50%), but not enough to push you above 50% for A (because you were already so far below 50%).
But the idea of meta control grouping seems to rule out "double standards" like this. What the idea of meta control grouping seems to imply is that, because you got 3 bits of evidence for A, and A is likely false, that means you can easily get 3 bits of evidence for a false proposition, and thus 3 bits of evidence is actually not evidence at all (or at least much less evidence than you thought). So what this would mean is, from now on, any time you get 3 bits of evidence for anything, you should either discount it, or "correct" it down to (say) 1 bit of evidence. For the sake of communication, I will call this idea Devaluation of Evidence (DE).
I can see this getting out of control very quickly. Any time you get evidence for a proposition you have low priors on, that amount of evidence becomes worthless for every other proposition you've ever believed. This should make you progressively more uncertain about everything over time, until (in the limit) every one of your beliefs is 50-50 and you have no beliefs about anything whatsoever. For the sake of ease of communication, I will call this idea the Progressive Devaluation of Evidence (PDE).
In fact, I almost want to say that this results in an infinite loop right off the bat. If 3 bits in favor of B needs to be corrected down to 1 bit in favor of B, then does that same 1 bit in favor of B need to be corrected down to 1/3 of a bit in favor of B, and then down to 1/9 of a bit in favor of B, and so on? I'm far less sure about this, but this seems to be a potential issue as well. (Let's call this the Devaluation of Evidence Infinite Loop [DEIL]).
In other words, unless I'm misunderstanding something, the idea of meta-control-grouping seems to break Bayesianism. I'm not certain about any of this, and I get the feeling I'm fundamentally misunderstanding something here, but I just wanted to get this out so that someone knowledgable can clear this up for me.
Wrote a short sci-fi story on AI and superintelligence.
Just 3 pages, enjoy!
https://docs.google.com/document/d/1fx_XNi2Nboqc4ztIh7IGQQHgT1sxpzVXEv8ux1XbRWg/edit?usp=sharing
Something I wish people would focus more attention on, when discussing "privilege" and "oppression", is the effect of habituation and the hedonic treadmill. After all, just because there's a relative difference between two people it doesn't automatically imply a difference in perceived satisfaction. People very quickly get habituated with their life situation, and as suicidal celebrities discover, having stuff doesn't automatically imply you're leading a happier life.
Are prices part of the map, or part of the territory?
I'm curious to what extent various social and psychological ills would turn out, upon closer and honest inspection, to be self-fulfilling prophecies. That is, the fact that people are made aware they should "feel bad" is actually the main cause of them experiencing negative emotions in the first place.
For one thing, I'm thinking about various "traumas" that people tend to get groomed into believing they've experienced, often in a cult-like manner. But I also have in mind the psychological effects of "inequality" (which are often the effects of becoming aware of, constantly perceiving, and obsessing over inequality, more than of the actual inequality itself).
How does one get their meetup link added to the spreadsheet? I must have missed any info on that... I have one for Ljubljana, Slovenia that I'd like to add somehow... https://www.lesswrong.com/events/G5gufXwuHhJuoDoLi/ljubljana-acx-meetups-everywhere-2021
Thanks.
The approved universities for the EA scholarship seem to be particularly elitism and classist for a community so aligned with Rationalist thought. I'd understand some combination of planned course of study, explanation, need and minimally prestigious university (which goes towards ability to change the world for the better). Instead it is limited to a particularly tiny set of schools. Maybe I'm making a molehill out of an anthill here, but this is the same flavoring that has put me off EA forums in the past.
Here is the Late and Great Hunter S. Thompson, 20 years ago today.
http://proxy.espn.com/espn/page2/story?id=1250751
I recommend reading the whole very short piece, the choice bits:
"We are going to punish somebody for this attack, but just who or what will be blown to smithereens for it is hard to say. Maybe Afghanistan, maybe Pakistan or Iraq, or possibly all three at once. Who knows? Not even the Generals in what remains of the Pentagon or the New York papers calling for WAR seem to know who did it or where to look for them."
"The towers are gone now, reduced to bloody rubble, along with all hopes for Peace in Our Time, in the United States or any other country. Make no mistake about it: We are At War now -- with somebody -- and we will stay At War with that mysterious Enemy for the rest of our lives."
Ouch.
I want to talk about Motivated Reasoning.
To recap, this is when you come up with a rational sounding reason for some opinion/action you want to defend.
I’ve long been aware of it, but I didn’t understand it is typically an unconscious process: Some optimizing part of your mind decides you want X. The Motivated Reasoning part then comes up with as good a sounding argument for why this is the just and right thing to happen. And then you - the tip of the iceberg conscious “you” - ACTUALLY BELIEVE IT!
This has some implications
Often people will argue using obviously broken or dishonest arguments. I used to think they were lying. But now I think they’re maybe BEING LIED TO, by their Motivated Reasoning Center.
This must also imply different methods of arguing with such people. The “come on, admit you’re lying” strategy is obviously out. What, if anything, are some better ways?
Hey all! I forget if this violates any rules about self-promotion on open threads (if so just let me know and I’ll delete) but I’ve recently been investigating how AI-generated art and text is draining away meaning. I think this is a big problem that people, especially the creative class, will be blindsided by. What happens when all music is generated by some non-conscious Chinese Room? I have no idea but I find it terrifying to consider. Any thoughts (or reassurances) are welcome https://erikhoel.substack.com/p/big-tech-is-replacing-human-artists
If a 10-year-old can spend a continuous 5 hours focused on building detailed environments in Minecraft, can said 10-year-old have ADHD?
Theodore Roosevelt and the Man in the Arena:
"It is not the critic who counts: not the man who points out how the strong man stumbles or where the doer of deeds could have done better."
https://newsletter.butwhatfor.com/p/the-man-in-the-arena-theodore-roosevelts