Ahhhh. Slightly more than comfortably full. Tryptophan addled. House a on the warm side of comfortable.Time to lie on the sofa in front of a football game I have no intention of paying attention to and uh uhhhnn….
A hypothetical situation/question that has bugged me since childhood: A ball is thrown at a wall. The distance between the ball and the wall can be measured and expressed mathematically. As the ball approaches the wall, that measurement becomes a smaller and smaller number. Numbers are infinite, and numbers can be always be divided and made smaller (in my understanding). My question - as the ball nears the wall, the numerically expressed distance can always be made into a smaller number - how does the ball, numerically speaking, ever actually make contact with the wall? This may be a dumb question, but hey, if you don't ask... Thanks in advance.
These are not trivial problems at all. If you suppose that the ball moves at constant speed of 1m/s toward the wall, and starts out 3m away, then the distance to the wall is (3-t)m and as the time t approaches 3, the distance goes to 0. There's nothing special happening at t=3 from a mathematical sense, though physically the ball contacts the wall at that point and you need to refine the model to go any further.
Actually, Newton developed calculus to deal with problems of the type you're describing. The way calculus is taught in schools and universities obscures this fact, but the basics of the subject are all about making sense of processes which happen in continuous time, like the one you describe.
You'll read this and think I haven't answered the question. Actually setting calculus on a completely rigorous mathematical footing took around 250 years, and the effort of many mathematicians. (Real numbers are already surprisingly difficult to describe formally - and then you're looking at processes involving these to describe the real world.) To get a completely satisfying answer to the question, you would probably need to go through a sequence of undergrad level calculus courses - but hopefully this gives a sense of what's going on.
Keep in mind that time works the same way; every time you remeasure the distance from the ball to the wall, you remeasure the time it takes to travel there. Eventually you hit... is it Planck, or Heisenberg? Either way, there's a distance at which things are no longer things, and then the ball changes direction.
Technically, the ball never touches the wall. The ball is made of atoms, the wall is made of atoms; if you put the atoms too close, they repel. At a tiny but nonzero distance the force becomes strong enough to make the ball bounce.
(If Chuck Norris throws a ball at a wall, a microscopic black hole is created at the place of collision, and then it gets complicated. EDIT: The atoms would probably just fit in places between other atoms and the ball would get stuck inside the wall, possibly leading to an explosion.)
You're looking for solutions for Xeno's paradoxes, which in the modern era broadly get solved with calculus. I'm not super up on the exact solutions, but it's something to do with limits?
Gebru and Torres, two dissident researchers who became famous as coauthors of the renowned "Stochastic Parrots" article, recently published a new paper critical of the rationality sphere and EA:
They coin the awesome-sounding acronym TESCREAL to describe the sphere, and then argue that it has roots in the eugenics tradition. Are you aware of this article, and have you considered responding to it?
In what way are they "dissidents"? They're representatives of the boring, narrow hegemony that rules intellectual life, trying to extend that hegemony into one of the few corners they don't control.
Once I strip aside the authors' tone of moral panic about the followers of TESCREAL beliefs (Transhumanism, Extopianism, Singularitarianism, Cosmism, Rationalism, EAism, and Longtermism) their arguments boil down to (a) these are all influenced by eugenics-based ideas, and (b) and most of their AGI hopes and dreams are not based on sound scientific nor engineering principles.
As for the first, they admit that the dream of breeding/creating superior humans goes back long before the early 20th-century Eugenics movement, and they fail to tie the current eugenical beliefs to those of the previous century. But they try to tar the current movements with old stains of the old movement.
And as for the latter, I tend to agree with their statements. A lot of the TESCREAL beliefs are attractive to smart people who think they know more than they do (and Charlie Stross has noted how much of their beliefs have been influenced by old science fiction tropes). I see the TESCREAL belief systems as being mostly harmless crackpottery. Of course, certain billionaire followers of TESCREAL beliefs could do a lot of social and political damage if they're given the power to do so.
Reading the article, commenting as I go, I am not promising that I will make it to the end:
> we have seen little discussion of why AGI is considered desirable by many in the field of AI, and whether this is a goal that should be pursued
Apparently no one reads Yudkowsky these days.
> In this paper, we ask: What ideologies are driving the race to attempt to build AGI?
In this comment, I answer: it's about money and power; the things that transcend all ideologies.
> the acronym “TESCREAL” denotes “transhumanism, Extropianism, singularitarianism, (modern) cosmism, Rationalism, Effective Altruism, and longtermism”
At the first sight, this sounds like a complicated way to say "nerds". (Later we will see how much the words can actually be substituted, but if yes, that would kinda reduce the entire finding to: "the people trying to build AGI are... nerds". Which is true, but not really surprising.)
> The idea of eugenics can be traced back to the origins of the Western intellectual tradition
Yeah, I guess if you want to talk about transhumanists and singularitarians, you need to start with the recent examples, such as Plato and Aristotle.
> second-wave eugenics arose in response to new technological possibilities associated with genetic engineering and biotechnology. Such technologies opened the door to human “improvements”
When the authors say "eugenics", they basically include anything that concerns human DNA. If you worry about your child potentially getting Tay–Sachs disease, you know who else worried about people having inferior genes, right?
The link between eugenics and transhumanism seems to be that eugenicists wanted to improve humans by improving the genetic pool, and transhumanists want to achieve the same goal by abandoning the biological limitations completely. Different methods, the same goal.
Extropianists were the first organized modern transhumanist group. Singularitarians are also transhumanists. This... apparently makes the belief that an artificial intelligence could be smarter than humans an eugenicist belief, by the power of association. (I wonder whether a belief that Google can find things online better than you is also a fundamentally eugenicist belief. Or that a calculator can do multiplication faster than you.)
Never heard about cosmism before, so whatever; I find it plausible that it is also somehow related to thinking about the future and technology (which, as we have already established, are fundamentally eugenicist topics). Rationalists... also believe that AI is a big thing, and that it would better for the poweful AI to be good rather than evil. (Those assholes.) Effective altruists want to alleviate poverty... I am sure there is something sinister behind that, too, let's wait for the punchline... ah, yeah, some of them also worry about the future humans. And longtermists, those by definition worry about the future. (Remember, worrying about the future is an eugenicist thing.)
> Indeed, transhumanism, Extropianism, singularitarianism, and cosmism are examples of second-wave eugenics, since all endorse the use of emerging technologies to radically “enhance” humanity and create a new “posthuman” species.
OK, still in 20% of the article, but seems like this is the crux. If you accept the premise that trying to improve humans in any way (such as curing hereditary diseases or augmenting humans by technology) is basically the same thing that Nazis tried to do, only using modern technology, then... either you become a Nazi sympathizer, or you need to ban all efforts to cure genetic diseases or augment humans. Anything else means being in denial; and luckily we have two smart professors here to expose your hypocrisy.
> The TESCREAL bundle shares certain “eschatological” (relating to “last things”) convictions. As with religions like Christianity, these take two forms: utopian and apocalyptic, which are inextricably bound up together.
The belief in either horrible or glorious future is also in Marxism, just saying. On the other hand, effective altruists are mostly worried about the dark present. (I think that rationalists are also quite concerned about people being stupid right now.)
> The reason for concern is that emerging technologies are expected to be (a) extremely powerful; (b) increasingly accessible to both state and nonstate actors; and (c) dual-use, as exemplified by CRISPR-Cas9, which could enable us to cure diseases but also synthesize designer pathogens unleashing an “engineered pandemic”.
Here I would like to ask the authors whether *they* agree with this assessment factually, or not. Because if they do, then it also puts them in the category of people worrying about the future. (And if they don't, then what's the point of obsessing whether the technologies are connected to the sinister TESCREALs?)
> in 1996, Yudkowsky expressed concerns about superintelligence, writing: “Superintelligent robots = Aryans, humans = Jews.
Ah, Yudkowsky being Yudkowsky. But what exactly are the authors trying to say? Are they suggesting that Yudkowsky proposes the extermination of Jews/humans as a desirable thing to do? Or are they just taking things out of context in order to make their point?
> “Intelligence,” typically understood as the property measured by IQ tests, matters greatly because of its instrumental value
Uhm... yes?
> The obsession with IQ can be traced back to first-wave eugenicists, who used IQ tests to identify “defectives” and the “feeble-minded.”
Ah, I see the logic. In a parallel universe where Nazis sent *short* people to death camps, it is probably a taboo to talk about human height, and if you suggest that putting things on the top shelf can make them difficult to reach for some people, you get cancelled.
A list of people and organizations to get cancelled.
> Effective accelerationists believe that the probability of a bad outcome due to AGI is very low
Yeah, that's pretty much the same thing as Yudkowsky says, except with "not" at the beginning.
(Would that make the acronym TESCREALEACC?)
> how TESCREAList groups are steering the field of AI toward the goal of creating AGI.
This feels like giving too much credit. Are they saying that without these people, it wouldn't occur to anyone to build an LLM?
> In 1955, four white men officially launched the field of AI
Ah, the field is forever tainted by the association with crackers. :(
> many researchers in fields currently associated with AI, such as natural language processing (NLP), machine learning (ML), and computer vision (CV), explicitly distanced themselves from the term “AI,” in part because it became associated with unfulfilled grandiose promises. Nonetheless, some groups continued to work toward “artificial general intelligence,”
So, both the former and the latter agree that the AIs we have now are not AGIs? And that is somehow controversial? Or why the framing of "some researchers work on X, but others [sinister music starts playing] work on Y"? Isn't it quite normal that different researchers work on different things? It seems like the authors are trying to hint at some dark controversy here, but to me it sounds about as sinister as saying "some biologists study mice, but others [sinister music starts playing] study frogs". Like, yeah, that's how research works. (Actually, I may be wrong here. That's how research works in natural sciences. In some other fields, it is probably very important that the researchers agree with each other and all do the same thing, otherwise they get cancelled?)
> How, then, would researchers know that they have achieved their goals of building AGI? They need to know how to define and measure “general intelligence.” Unsurprisingly, these definitions rest on notions of “intelligence” that depend on IQ and other racist concepts
(facepalm)
> For these reasons, Keira Havens, who has written extensively on race science, asks those attempting to build AGI: “Why are you relying on eugenic definitions, eugenic concepts, eugenic thinking to inform your work? Why [...] do you want to enshrine these static and limited ways of thinking about humanity and intelligence?”
Mussolini promised to make trains run on time. Why are you still using trains?
> One conjecture is that the resulting AGI will be so intelligent that it will figure out what the best thing to do is in any potential situation.
I'm confused. So did they read Yudkowsky or didn't they? Or were they just looking for a quote to take out of context, and ignored everything else, including the part where he *literally* called exactly this thing "the sheer folly of callow youth"?
> TESCREALists have been able to divert resources toward trying to build AGI and stopping their version of an apocalypse in the far future, while dissuading the public from scrutinizing the actual harms that they cause in their attempts to build AGI.
Uhm... sorry about mentioning this Yudkowsky guy all the time, but you included the movement he started in your acronym... and you pretend to be experts on what the movement is about, so... maybe it would be nice to get familiar at least with the very basics?
> As an example, one of us worked as a hardware engineer designing audio circuitry for devices such as laptops. Some of the tests that we performed as part of our work included drop testing, constantly dropping devices to understand the manner in which their functionality degrades when they are exposed to shocks, placing the devices in extremely cold or hot environments, frequently restarting them [...] we argue that the first steps in making any AI system safe are attempting to build well-scoped and well-defined systems like those described as “narrow AI,” rather than a machine that can supposedly do any task under any circumstance.
OK, I really don't want to put your expertise in restarting laptops in doubt. But, you know, these AI things sometimes turn out to work differently than you would expect. For example, I think that scientists in 1960s believed that we will make artificial intelligences understand human knowledge one topic at a time, by building expert systems and cleverly designing the semantical nets. They would probably be shocked to hear that the state-of-art AIs of 2024 are good at poetry and painting, but suck at math. And yet, here we are. The line between the narrow AIs and AGIs may turn out to be similarly unpredictable; it can happen much later or much sooner that we expect, and may be caused by an improvement that we will not suspect of having exactly that impact.
(THE END)
OK, my summary would be that the authors identified a group of people whom I would describe as "nerds thinking about the future, specifically the technology", and correctly noticed that there is an overlap between various groups.
The connection to eugenics is mostly made through "trying to improve things" and "believing that intelligence is a thing". I understand that both of these things are *problematic* for some people, but I also think those people are... how to put it politely... not thinking clearly, and probably not even interested in thinking clearly.
Some of their arguments require *a lot* of cherry-picking. If you fail to notice that the effective altruists mostly care about reducing the suffering that exists *today*... or if you believe that Yudkowsky wants to build the AGI as soon as possible, and believes that by the virtue of being generally intelligent it will get all the answers right... then I strongly doubt your expertise on the individual elements of the "TESCREAL" boogeyperson you invented. It's more like you sometimes need to make things the very opposite of what they are, in order to make your point.
Congratulation on finding your powerful and rich enemy responsible for manipulating the society and science towards their sinister goals. You know who else did that, right?
Yes. As usual, it is 10% of reality that gets 90% of the attention, but see e.g. here: https://www.givewell.org/charities/top-charities -- malaria, malaria, vitamin A, some unspecified vaccines. Okay, I guess the vaccines are about reducing suffering in the future, but it is still different from what the "TEASCREAL" model suggests.
Maybe somewhere lower on the list there is a longtermist project, and that is what makes some people freak out. Or maybe there are also separate lists by categories, and the longtermist category contains longtermist projects. But there is a difference between "there are some X somewhere among Y" and "Y is about X".
I coin a slightly less awesome acronym HGT, which means Hitler, Gebru, and Torres. I claim that HGT has roots in Nazi traditions, and I have historical data to prove that at least one third of HGT was actively involved. That sounds quite damning, in my opinion.
Long story short, you can choose a perspective that supports your conclusions. As a rationalist, you should also reflect on how much your conclusion reflects the territory and how much it merely reflects the perspective you chose. To get the best results, you should aim to "carve reality at its joints", i.e. let yourself be guided by how things actually are, rather than your arbitrary definitions, no matter how clever they may sound (they may or may not be correct). For example, I am pretty sure you can find a connection between "TESCREAL" and eugenics tradition, just like you can find a connection between the eugenics tradition and progressive politics. The question is, is the former connection stronger than the latter? (If not, why do you focus on it so much? Instead of e.g. writing an article about "TESCREAL" having roots in the progressive politics?)
...okay, now I am going to actually read the article.
EDIT: One of the names sounded vaguely familiar, took me a moment to remember why:
Still haven't read the article, but now it seems to me that the alleged main connection between eugenics and "TESCREAL" is... consequentialist thinking, and worrying about the future of humanity. Technically true. (So it's like, someone says: "I would prefer if AI didn't kill all humans", and the response is: "so what you're saying is that you are worried about preserving the future of the while race?")
Well I skimmed through it... and the paper doesn't seem that wrong at the object level. Hell, even the last survey Scott posted did show that the readership for the blog is relatively pro-eugenics (but only when he didn't use the word "eugenics"). It's just that the authors' values are completely idiotic. What is with these people's obsession with identity politics and "marginalized communities"? Calling people "racist" just for trying to improve humanity is just... well, it's certainly proving their opposition's point that all of this is necessary.
I agree. Unfortunately, these values are dominant in Western academia and their supremacy is jealously guarded. Some high-profile EA should probably make a response, otherwise I'm worried that good EA ideas will see even more opposition from the academy.
Well that shouldn't matter, academia is going to get ideologically purged soon anyways. Though, I guess we'd have bigger things to worry about at that point...
By the new administration. Do you think the Republican party sacrificed all of their values just so they could maintain the status quo? The people, and the people that lead them, want change, and the only way to accomplish that is by rooting out the rot at the top. Not just in academia, but in corporations as well.
Of course, they almost certainly hate EA as well for being filled with leftist vegans, so uh... Good luck with that.
The Trump administration last time around couldn't even manage to purge the Trump administration. This time around I think they'll be ever so slightly more effective, but I doubt they'll manage a purge of academia.
Is the rationalist movement privileged? Ok, stupid question. I'd say the answer is an obvious yes. What I mean specifically is privileged in the economic sense. I.e. something that some wealthy people can afford to do/be but less wealthy people can't afford to be/do. And when I say rationalist I mean someone who esteems truth for truth's sake. Feel free to argue for or against this first question, but I take it as a given. It seems obvious to me that some people are getting subsidized by their jobs/career to be rationalists while most other people get subsidized to shut their mouths and do what they are told. So my actual question is, in what ways can one incorporate rationalist principles in their life if honing these skills may actually end up hurting them economically? Do you see it sort of like a Socratic matter of principles (I'll be broke as hell but at least I'll know I was being honest with myself), is the world really just dog-eat-dog and poor people should just excise rationalist thinking from their lives if they know what is best for them? Other ways of looking at this? I ask because I work as a secondary ed teacher and I've found that there are aspects of my job where being a rationalist is helpful/good but I think it is valuable to bifurcate rationalist thinking (truth for truth's sake) and just play the political game a lot/most of the time.
Good question. Why not take a data centred approach?
How do we measure privilege? It's going to depend on the society you're immersed in, and your own values. But most people would agree that wealth, education, a 'good' job, social standing and various freedoms/civil rights are a reasonable measure of this. The first three are fairly easy to measure. In the amount of time I'm willing to spend on this, I can compare the ACX survey to US demographic data.
Something like 85% of Scott's readers have a third level degree, upwards of 20% have a doctorate. Less than 40% of the population in the US has a bachelors degree. Approx. 2% have a PhD and at most 5% have doctoral level qualifications. So readers of the blog do have educational privilege. Since the ability to do a PhD requires years of forgone earnings, it's reasonable to infer that blog readers are coming from a family background which is higher socio-economic status.
Unfortunately, the survey data just reports numbers for income with the majority of responses censored. You could ask Scott for the data if you wanted to do more analysis, but I think it's likely that average and median earnings are well above the US average. Participants did self-identify their SES but I would treat this with caution, it's better to collect 'hard' data rather than self-descriptions, particularly when something like 'upper-middle class' is context and country dependent. Personally, I think the level of education and amount of free time required to read blogs like this and post in them means that they're biased to the privileged.
Recently, I've been at odds with my manager over small issues. E.g. they wanted to rearrange the schedule of a meeting I was chairing so they could arrive and deliver the welcome address 30 minutes in - I insisted the welcome be at the beginning, they could speak on other topics if arriving at the end of the meeting - they rescheduled to arrive on time. I don't think this is in any way a rationalist approach - to me it's common sense. I had the background and sufficient social standing to make my case persuasively. I didn't play the political game though - and I'm lucky to have a boss who is (mostly) willing to listen to pushback. I'd be curious to hear if a card carrying rationalist would have approached this differently - to me it was a mixture of common sense and my own stubbornness.
Not especially; the non privileged rationalist don't call themselves rationalists because that term is intensely weighed down (unfairly) in the consensus by your Theilites and your SBFs and your weird cryptofascist orbiters.
Eg, they decided not to pursue lots of money for moral reasons; they are out there digging wells or being teachers or doing the type of research that doesn't end in you being super rich for inventing the Juicero.
I would say that it's hard not to be a rationalist if you are at all introspective.
A) The Big Yud often said that "rationality is the art of winning". I.e. winning is the goal, rationality is a tool. If rationality precludes you from winning, discard rationality and win. He typically relies on Zen motifs to get his point across. C.f. "The Twelfth Virtue" [0] and "Something to Protect" [1].
B) The wealthy are privileged ipso facto. Dependency (of any kind, including financial dependency) represents exogenous influence. Unfortunately, specialization has made modern society quite interdependent. So it often is, infact, in your best interest to adeptly navigate Baudrillard's Simulacrum [2]. Another way of thinking about money: money is "slack" [3], and slack is options. Including the option to be intellectually-honest.
> So my actual question is, in what ways can one incorporate rationalist principles in their life if honing these skills may actually end up hurting them economically?
C) To paraphrase The-Bugman-Who-Must-Not-Be-Named: "you have to live like you're in enemy territory". If the powers-that-be don't like what you're teaching your students, you need to realize that you're basically distributing samizdat.
Hm... I don't remember anything like that. Got links? Not gonna lie, it's been a long time since I read The Sequences. And I never completed the entire corpus. So I'm prepared to be surprised.
Whatever the case, I stand by my initial comment, insofar as winning is ultimately more important than truth. (No, this isn't a carte blanche to act like a pathological liar. But Kant's radical honesty, for example, is a kinda bonkers imho.)
Movements are not privileged; people are. Some rationalists are rich, some are poor, most of them probably somewhere in the middle. Or slightly above the average, considering their typical professions.
> something that some wealthy people can afford to do/be but less wealthy people can't afford to be/do.
My personal freedom of speech is mostly a result of living in Eastern Europe where no one gives a fuck about what you think. If I lived in America, I would probably watch my tongue more carefully. I am *not* wealthy enough to not need a job, which dramatically limits the things I can do (and in America would also limit the things I can say).
> It seems obvious to me that some people are getting subsidized by their jobs/career to be rationalists while most other people get subsidized to shut their mouths and do what they are told.
Yeah, your type of job determines what kind of idiots you are surrounded by.
> Other ways of looking at this?
It seems like some good (for self-preservation) habits from communism are also useful in the woke era, so here is one:
* Don't think.
* If you think, then don't speak.
* If you think and speak, then don't write.
* If you think, speak and write, then don't sign.
* If you think, speak, write and sign, then don't be surprised.
Personally, I would think, and only talk to friends or talk pseudonymously on the web.
> I work as a secondary ed teacher and I've found that there are aspects of my job where being a rationalist is helpful/good but I think it is valuable to bifurcate rationalist thinking (truth for truth's sake) and just play the political game a lot/most of the time.
Yeah, you can use some rationality at your work, without saying explicitly that this is what you are doing. (When someone important says "there is a dragon in a garage", say "of course there is", and then continue using the garage as usual. Which is what many people actually do.)
Others may chime in with different takes, but, for me, rationalism is about forming confidence estimates related to beliefs, it is not about saintly purity in how you choose to act or what you choose to say.
Taking an example from inside your question: imagine a situation where you are called to make a statement and you see two choices (a) what you believe to be true, (b) what you believe to be politically expedient. Rationalist methods could help you in the following ways:
(i) identify a false dichotomy and determine a third (or even more than one) alternative thing to say/way to act
(ii) form and update an estimate of the probability that (a) is a better/worse choice than (b)
(iii) phrase your belief (a) in a way that does not have the assumed political cost (because you indicate openness to other possibilities, you support your statement with evidence, you shade your statement probabalistically)
Last year I found myself across the table from Substack CEO Chris Best—and I hoped he would satisfy my curiosity on one matter. So I asked:
“Chris, I have a question. Has Elon Musk made an offer to buy Substack?”
Then I added:
“Appropriate answers to this question are (1) Yes, (2) No, or (3) I prefer not to say.”
Chris didn’t answer immediately. He thought it over for a moment, then responded: “I prefer not to say.”
In this soft-spoken way, Mr. Best showed his wisdom. Although he works hard to empower journalists, he’s smart enough to know to watch his words around those pesky folks. We’re as leaky as sieves, and only half as reliable.
But the New York Times reported yesterday that Elon Musk did attempt to buy Substack last year. Musk also hinted that he would merge it into Twitter and let Best run the combined companies.
The Times also reports that Best rejected this offer.
And he didn’t even fret about it. It was a “short-lived discussion,” according to the Times’ three sources.
That proves that Chris not only has wisdom, but also possesses integrity and strong core values. When he says that he supports indie writers, he really means it.
If Musk’s plan had been accepted, Chris would have enjoyed a huge payday in the transaction, and also might have found himself running the most powerful media force in the world right now.
But he didn’t hesitate to walk away from that.
I want to commend Chris and the entire Substack team. They not only talk the talk, but walk the walk—which is rare in the media world right now. Or in any world right now.
First, saying "I prefer not to say" is not necessarily the same as kinda-deniable "yes". Maybe he rejected Musk. Maybe he is waiting and hoping that Musk will offer more money. Maybe someone else offered money, and he is trying to make the other person think that they are competing against Musk. Maybe he even said yes to Musk, but they have some kind of deal that the website will officially change hands on January 1st 2025, and until then there is an embargo on this information.
From business perspective, the Substack model seems great. By taking a fixed fraction of subscriptions, the financial interests seem aligned; Substack makes money when the authors do. The possible downside is that this can make the platform attractive to all kinds of conspiracy theorists and scammers (I imagine that writing Flat Earth News could be quite profitable), but if you advertise yourself as a platform where no one gets censored, that already goes with the territory.
But technically, using Substack can be quite frustrating experience, and frankly it seems like it's no one's priority to improve it. It took years just to make ACX comment section usable, and it still could be improved a lot. Or the bug where if you replied to a comment in an e-mail, your reply got inserted incorrectly in the comment tree. (That one is probably fixed, but that also took years.) I haven't written a post lately, but I remember it to be also frustrating. Like, how am I supposed to write articles containing code examples, if all existing character style options are so huge that you can barely fit 40 characters in a line? Or why do I have to choose whether a word is monospace or a hyperlink, but it cannot be both? It's obvious that the intended type of article is "type a lot of text, drag and drop a few pictures and a YouTube video, insert five 'subscribe' buttons, and call it a day", which I guess is where most money comes from. Yeah, it makes sense economically, but that means that if your technical skills are even slightly above zero, Substack will constantly make you cry.
So, maybe don't sell to Musk, but at least hire someone who is technically competent and who cares.
Are we sure Musk didn't just lowball him? There's no practical reason for him to overpay for Substack like he did with Twitter, considering that this site is already pretty right-leaning.
And reading the article more closesly, Substack still isn't doing very well financially... Musk has more than just money as leverage, given that he immediately retaliated by hiding posts with substack links, not to mention his new political connections. He'll get what he wants eventually.
Not that I think that'll be a massive issue anyways; all that'll happen is that everyone at Substack gets fired and the site gets merged into X. And also he'll probably stop renewing most of the contracts with the writers here. I guess that'll suck for Scott.
One of the weird things happening in the amazing nowadays is the number of people eager to jump to defend Musk. Like, you know he's a man with immense wealth and power, and is quite capable to looking after himself. He's also not running for any popularity contest. So... why do so many spend so much effort defending whatever he does? Jumping to counter every lightest slight in his direction, like he was a precious baby or something...
That's weird, because I got the opposite impression, that recently it is incredibly popular to accuse Musk of... well, nothing specific, just being generally the worst. Seems to be so since the moment he bought Twitter, but maybe I got the timing wrong.
I don't even like the guy; I think he is neither a saint nor a devil, just a smart rich guy trying to become even more rich, who seems to spend too much time online shitposting.
Is there somewhere a list of his "crimes" that I could read?
For the same reason that lots of people jump to criticize him for no real reason. He's become a symbol of the culture wars, so everything he does gets framed that way.
Also I don't think it's fair to characterize my comment as "jumping to counter every lightest slight". I genuinely don't think him buying substack would be bad for anyone and don't understand why someone would reflexively assume that it would be.
He makes rules on the fly and interprets hazy policies in a way that justifies him banning, suspending or pushing voices he doesn't like off the platform. When he took over I was curious to see what twitter with free speech would look like, but unfortunately these are not the actions of someone with a genuine commitment to free speech.
Wait! Wait! The Webb Telescope disproved the existence of Dark Matter?! I guess I must have been snoozing. The massive galaxies that Webb observed in the early universe are an excellent fit for MOND, and their existence that early in the universe pretty thoroughly falsifies the Dark Matter models of galaxy formation.
Stupid question: if Modified Newtonian Dynamics (MOND) is a better description of gravity, how does this affect the value of the Cosmological Constant? Does Dark Energy also go away? I don't know enough about MOND because I never took it seriously. I guess another Internet rabbit hole awaits my exploration.
Hmm, well I've been following Mond for a while now. Tracy McGaugh's blog is full of stuff. https://tritonstation.com/
Later edit;
I just want to add that Mond is less of a 'theory' and more of a 'fit to the data'. If gravity at accelerations below a_0 became a 1/r force law, (rather than 1/r^2) then that fits all the data.. up to some point. Mond fails at the mass level of galaxy clusters, it doesn't fit the third hump in the CMB data. nor the bullet cluster... There's something missing from Mond.... My favorite idea is some hot dark matter... sterile neutrinos. (I mean first of all I love the name.) Oh and that Mond is true... whatever mond is... beyond a fit to the data.
I don't think it can be concluded yet that DM or that the age of the universe has been falsified nor that has MOND been proven.
On this topic, the empirical evidence for MOND seems better but that isn't a 100% exclusion for DM.
DM can explain current star formation history, if and only if models of star/galaxy formation "innovate".
Be it a non constant IMF, super eddington accretions, assumptions about black holes and dust prevalences and properties, etc.
We are at the limit where nearly 100% of baryons needs to be converted into stars which sounds highly unlikely though. There has been some pre JWST era models that match most current observations. Moreover there are many possible more exotic catalysis mechanisms of galaxy and star formation, among which some might require new physics or standard physics but in practice non standard physics (few people will understand what I mean here), e.g. via dust based Mock gravity, rydberg matter or Zeldovich radiation.
Most importantly the debate should not be dumbed down to MOND or DM, one both can coexist, two there are many extensions (interacting DM, etc), three both can be false, the true theory might be quantized inertia, etc
My personal belief is that the age of the universe is at least moderately wrong.
The MOND existence will be forever settled via GAIA DR4 in 2026 (is settled in 2025 but ESA will gatekeep the data because they do not follow open data science)
BTW here is the biggest issue in LCDM by far, and you'll never hear of it, like in most sciences, the most interesting discoveries are absolutely unknown because well there is only one person on this planet reading the scientific literature extensively (me)
In reference to your paper, how would a missing dust signature in the CMB support the existence of dark matter? Would the presumed distribution of dark matter—clumped around galaxies and diffuse to a near-perfect vacuum in intergalactic space—reduce the density of dust in intergalactic space?
This is Scott's substack, and he has the power and the right to draw whatever lines he wishes. But in the past, Scott used to post a separate political open thread every week (month?) that allowed political discussions and arguments. He did a good job snuffing out the political crap on the other threads—but I suspect it became too onerous for him. And so, I'm forced to wade through political crap to get to any interesting philosophical or scientific discussions. This is one of the reasons I haven't re-upped my subscription to this sub.
Matt Gaetz has already withdrawn from the Attorney General nomination, having previously resigned from the House, meaning he's currently not in any political offices. This removal of Matt Gaetz from Congress has the chance to be the first win of the new Trump cabinet, although I don't know if his resignation will actually hold this early in.
It aeeems that nobody who writes about these things in the news and elsewhere is familiar with negotiating or what they call in VEEP "cock-thumb" it was immediately obvious to me that that's what trump was doing with Gaetz.
Put aside the question of whether Sanders is right or wrong at the object level - my opinion is that Israel has a right to self-defense, crushing Hamas/Hezbollah is in the interests of all non-insane people, and while all conflicts have occasional outrages the IDF has overall done a good job in a challenging urban environment - the war is divisive in America. Whether those opposed are right or wrong isn't the question. Everyone has a right to their opinion and proportional say in how American power / tax money are used, dumb though their opinion may be.
What's a rough percentage of Americans opposed to this or some policy generally that you feel is sufficient opposition to be placated? Does it need to be a majority? 50%, 25%, 10%, etc.?
Would it be smart politically for the Democrats to cater to that opposition? For Trump? Would it be unifying or equally divisive for the government to say that the cost in domestic dispute is too high and Israel should source weapons elsewhere?
Democrats basically alternate between cheerleaders for the military (who quickly fall out of favor when wars go badly) and diehard pacifists (who get credibly accused of being America-hating foreign agents).
It's smart to have both wings so that a candidate from either side can be put forward when the situation calls for it. In previous conflicts, the pacifists have been either really, really right (Iraq the second time, e.g.) or really, really wrong (Iraq the first time, e.g.), so it helps to have a wide range of stances.
Republicans do the same - Donald Trump represents his party's isolationist wing, whereas George W. Bush (the previous Republican president) represents the interventionist wing.
Hillary Clinton (2016 nominee) voted for the Iraq War. Barack Obama (2008 President) voted against it.
Meanwhile, Barack Obama campaigned on the idea that Afghanistan was the "good war". His Vice President, Joe Biden (2020 President) withdrew from Afghanistan, letting it fall to the Taliban.
So the Democrats liked the Iraq War enough to vote for it, and to vote against it. They liked the Afghanistan War enough to campaign on winning it, and then to lose it.
I think it's easier to describe that as two separate ideological wings living uncomfortably with each other, rather than one group exercising a consistent ideology.
Wasn't Biden left with a fait accompli by Trump who'd made continued occupation untenable by the time he'd left office (by making a deal with the Taliban, letting the Taliban infiltrate governmental zones, and making physical preparations for total withdrawal)?
That's one way to frame it. Of course, when he was running for President, he discussed withdrawing from Afghanistan - look at the 2020 Democratic Party Platform and you can read their plans of Afghanistan under the heading "Ending Forever Wars". It's pretty much identical to Nixon's peace with honor that ended the Vietnam War.
Contrast that with 2008 - the chapter on Afghanistan is called "Win in Afghanistan", with sentences like "We will send at least two additional combat brigades to Afghanistan, and use this commitment to seek greater contributions–with fewer restrictions–from our NATO allies."
So I find it hard to believe that Biden wanted to surge additional troops into Afghanistan, a la Obama, but was thwarted by Trump. I think it's easier to believe that both Biden and Trump wanted to end the war in Afghanistan, but blame the other party for the inevitable bad press.
There's 13 years between the 2008 campaign and the 2021 withdrawal; plenty of time to change your mind based on the "facts on the ground" (a rare actual military use of the cliche). Also, two different people.
I'm trying to argue that the Democrats have two different wings - one pacificist/isolationist, and the other interventionist. I'm comparing the most recent Democratic President with the previous Democratic President - those two are unfortunately separated by 12 years. That's an unfortunate, but unavoidable part of the data set I'm working with.
You could however look at a single slice in time, with a single chamber. Look at the 2003 vote on the Iraq War - 81 Democrats voted in favor, 126 voted against it. For comparison, 215 Republicans voted for it and 6 against. The Republicans at the time were very much interventionists. The Democrats? They were split - not perfectly split, but a 60-40 split is a very real split.
That's a fair point on my wording. I apologize. It was a quickly written post and I shouldn't have used that framing. I was thinking more of Hamas in light of Oct. 7th but it was much too forceful and broad. Consider it withdrawn.
If you punch me in the face, don't complain when I rain punches on you. "But you're the one punching now! You ought to stop!"
Israel's objective is not to produce refugee camps or mass graves, but to prevent being attacked again. Were they trying to wipe out the Palestinians, they're doing a bad job given their resources. They are showing as much restraint as reasonable in limiting civilian casualties.
I don't expect much support on account of society being composed of shortsighted stay-the-course cowards who reserve their strongest antipathies for outsiders and who will die defending their leadership absolutely no matter how bad they are. I assume this comment wil go over like a like a lead balloon thus proving my point.
Even the people in Room 101 loathed Emmanual Goldstein.
The first paragraph not so much. It's perfectly possible for a response to be excessive. If someone punches me once, and then I "rain punches down on him," continuing to just beat him until well after he's been turned into an unconscious broken piece of meat, I'm well past the point of a reasonable self-defense and should expect others to intervene, and possibly to be convicted of a crime. A lesser crime than I would be if I was unprovoked, but "he hit me so anything goes" is not the relevant standard. At a certain point, when "you're the one punching now," you really *ought* to stop.
> I'm well past the point of a reasonable self-defense and should expect others to intervene, and possibly to be convicted of a crime
I'm sure you can get away with that if you have the blessing of the US government and military. The justice system exists to serve a need that isn't applicable in military conflicts, or any other state-sanctioned violence for that matter.
Well, it wasn't a perfect analogy, as none is. But you're right from one standpoint: one punch doesn't give license to punch forever. My point was more that you can't just call to stop just because you're losing.
And Israel, in this case, is trying to prevent the "first" punch from happening again. I do believe that if they were convinced it would stop, they would cease hostilities. But they think, justifiably, that if they stop it will just give their enemy a breather to prepare for another devastating attack. And then what is the proper response to prevent future attacks? Genocide?
what is the proper response to a conflict that was started by their forefathers but then they turned it up to 11 since ... let's say 1995?
they tried to ignore the simple fact that their co-tenants are not happy with the arrangement they forced and continue to force on them.
of course - turns out! would you believe it? - that after decades of marginalization said co-tenant's behavior shows, rather concerningly, heightened aggression.
and as a complete surprise, the hardliners continued promises that they can keep the peace ... failed.
....
yes, this is a big fucking Situation (waaay beyond Troubles level), with heavy military aspects, so the use of military force is not unfortunately some kind of "crossed line".
the moral arguments usually focus on the cruelty (and in some sense futility) of what the IDF does, and how it does it, not that it has "the right"
not to mention that - as far as I understand - the concept of self-defense is not well defined or really applicable for states occupying territories
so, on one hand, the IDF could go slower, be more careful, the tunnels won't run away after all. the only semi-sane objective is long-term occupation, otherwise the tunnels will get built again
on the other hand this is just more of the same, this guarantees more instances of mistreatment. (and even typing the words international peacekeeping mission led a laugh track fading in from somewhere.)
... and all that said, what is the proper response to being in this fucking terrible pickle? well, a bit more brotherly love would go a long way, but apparently there's been some mixup in the warehouse and thus recently folks are mostly getting war, far-right authoritarians and mayhem in other shape and form.
Most likely it has to start with showing self-restraint. (While not neglecting this elusive self-defense either.) But as long as Israel cannot really commit to dealing with its own fundies it will have an ample supply of counter-fundies :(
Agreed on both counts. Though when to stop is a more complex consideration than if you happen to be the one happening to be punching at the moment.
I'd expand a little on the second paragraph. The delta in mass graves / refugee camps between Israel and Hamas don't - as far as I can tell based on Hamas' stated goals and actions - reflect a difference in intents, only capabilities.
Ark is leaving out the part where he got punched in the face a while before the current punch... and punched again a while before that... and before that... and pretty much on a regular basis going back several decades, including a few rains of punches here or there.
In other words, the current punch isn't the only punch; it is the latest, and delivered as if there will be more, and maybe even a rain if permitted. That's why he rains punches back.
It's even worse than that. A few punches ago, he punched back more than once and "you're the one punching now" *was* the argument used to get him to stop... and then he got punched again. So that argument fails twice as hard now.
No, it's really not. It's a problem for a bunch of loud yutzes, but they poll in the 5% range, and literally no one else cares except for the people (>10%) who are wondering why Israel hasn't been tough enough.
>Everyone has a right to their opinion and proportional say in how American power / tax money are used
Yeah, no. Literally no one gets a proportional say in how their tax dollars are spent, most especially including the yutzes. I'm not sure why you think this either is or should be the case, but no, that's not happening.
The right answer for Dems is just what they've been doing, which is to pretend the yutzes don't exist and that they don't have to do anything about them.
> Yeah, no. Literally no one gets a proportional say in how their tax dollars are spent, most especially including the yutzes. I'm not sure why you think this either is or should be the case, but no, that's not happening.
You're correct that we, wisely, don't take an Athenian style vote on major decisions or budget line items. However, everyone does get an equal vote when it comes to elections - theoretically equal, state level distortions for the Presidency not withstanding, at a minimum equal at the level of your state.
Also agree with Peter Defeel below. If you're American you have ~335 million brothers and sisters. There are always going to be significant portions that see you as a yutze. You will be a yutze to your future self and vice versa. But there are no correctness requirements for voting. We don't throw out the votes of someone motivated by fear of subterranean lizardmen or equivalent. Like it or not, everyone with the right to vote gets a say even if their reasons are objectively or, to you, deeply stupid.
The question is also more of about the "ought" versus the "is". Yes, an administration with sufficient support from elected officials has the formal power to do whatever it can. But imagine a scenario where Examplestan is in an armed conflict with New Somewhereville. Magically reliably polling shows 50.1% in favor, 49.9% against. To what degree as a matter of national unity ought the views of a minority position be considered?
Well, that's why the two-party system was nice: it made sure the populace didn't get to have a voice in regards to actually important matters. Of course, the Republicans ended up defecting, but a one-party system effectively does the same thing.
Almost everyone who supports a "ceasefire" is imagining a particular set of terms for that ceasefire. Which they imagine to be obvious and achievable, so that as long as the war continues they'll say "stop it, you guys, cease fire now! [on my terms]". If the firing ceases and their terms are not in place, they'll say "that's not what I meant and you know it!". It's win-win for feeling good about yourself and blaming the bad stuff on other people.
So, what are the terms for the cease fire you're implicitly advocating, and what's the line for "that's not what I meant by ceasfire"?
It is impossible for an electorate to have in-depth, technical views about how things should be done. Instead, it is ideal for them to have an understanding of what they want done and for the political leadership to figure out how to reconcile their desire with the hard realities of the matter. The pro-Palestine people want peace and the upholding of international law. The US could work towards these aims by leaning on Israel a lot harder than they have been doing and forcing Israel to restrain itself when it comes to damage done to civilians, journalists, and other noncombatants. Israel could otherwise pursue its ambition to eradicate Hamas under these constraints.
That's awfully vague as a response to "what are the terms of the ceasefire you are advocating". But I notice that, at the end, you've got Israel continuing to try to eradicate Hamas.
In which case, what you are proposing is almost definitionally *not* a cease-fire. Unless you have a serious plan to eradicate Hamas without shooting at them, or unless you weren't serious about Israel continuing to try to eradicate Hamas.
Hamas delenda est; all else is commentary. If somewhere in that commentary is a serious, actionable plan to delenda Hamas with fewer civilian casualties, great, but all I'm seeing is a wish or a command for such a plan to exist.
It would be expedient for police in Europe to carry firearms everywhere they go and to shoot first and ask questions later like their American counterparts, but it has been deemed by democratic political processes undesirable that expedience should take such precedence over other concerns, like the possibility of innocents getting shot. Just because eradicating Hamas would be made easier by ignoring the international community doesn't mean that it is desirable from a broader perspective or an ethical one. America is sacrificing its own values by supporting Israel, as it is going against the very rules-based global order it created, currently manages, and has always championed, where nations are supposed to submit themselves to international consideration rather than act unilaterally out of realpolitik and pure power-consideration. America is torching its own moral authority. Even from a pure realpolitik perspective this seems to be more damaging to America than beneficial.
The practical limits Israel would need to adhere to would be no greater than those America accepts when waging its foreign wars. It is surely still possible, but an effort towards optics would have to be made, a fig leaf adorned.
A quick search of recent polling didn’t give me any answers to these questions but suggests younger Americans are less enthusiastic about military aid to Israel than older ones.
I personally like to see Senators fostering debate on big foreign policy issues and would like to see it happen more often, even if it doesn’t lead to legislation.
Sanders is 83. He ran in the 2016 and 2020 Democratic primaries. Both times, it was clear the Democratic establishment was against him winning and both times Sanders ultimately failed. Any hope he has of one day being US President has to be a massive longshot at this point. People know that he doesn't speak for the Democratic party as a whole.
So, he might as well just voice his conscience, his honest opinions on things.
He never actually joined the Democratic Party, still hasn't. It was odd that a sitting legislator who'd pointedly declined to be a member of a political party for decades, could enter himself into primaries of that party. (Will Rogers' Depression-era joke was "I don't belong to an organized political party -- I'm a Democrat.")
I suppose by letting a non-party member run in their primaries and seeing him get some substantial votes in them, the Dems are then basically stuck with him as far as public impressions go.
Politicians should, as clearly as possible, say ahead of elections what they will do if elected, and in which cases they will exercise their judgement in new situations and based on what.
Bernie sanders is acting predictably. So is Trump. So are most politicians, from the perspective of their voters, on this issue.
Agreed. Voting is an extremely lossy signal though. A strong isolationist, America First style Trump voter may be completely opposed to all military aid but still vote for him although he says that he'll increase support to some countries.
>Would it be smart politically for the Democrats to cater to that opposition?
Yes. I assume Republicans already get the voters who really want to support Israel. Democrats should offer the opposite option for voters who really don't want to support Israel. Not doing so leaves an opening for Republicans to take at least some of those votes. Eg, Trump did a lot of outreach to Arab Americans before election day.
Similarly there should be parties that stand for isolationism so voters can express that preference, which is why I think it's a good thing that neocons have been pushed away by the Republicans as the Democrats have increasingly embraced them. I also don't like the chattering classes trying to set the agenda by making certain preferences verboten, whether it's restricting immigration or asylum (increasingly less so today), being against support for Ukraine, etc.
That seems odd. Is there a minimum threshold of votes before a party should take an opposite approach to grab those voters? Pro-murderism seems open, for instance, but seems like it would do more harm than good.
The Republicans absolutely do not get all of the pro-Israel votes, and the Democrats are pretty evenly divided on the issue (with older Democrats overwhelmingly in favor of Israel while younger ones are only somewhat disapproving). That's a lot of voters to lose, and pretty much doom for the Democrats if they did. Supporting Hamas over Israel is one of the biggest losers in electoral politics, similar to things that nobody would consider part of electoral politics (even if not as bad as being pro-murder).
>Is there a minimum threshold of votes before a party should take an opposite approach to grab those voters?
Yeah, it's about 5%. If you try to grab single-issue voters below 5%, you'll end up losing more votes from either people who quietly disagree with that 5%, or who think that anyone who would go after that 5% is a sleazy charlatan. The Ds somehow realized that anyone who backed Hamas would be thought of as either a lunatic or a sleazy charlatan, and miraculously avoided doing so. I know! I'm as surprised as you are!
Where is this magic 5% coming from. This seems to be arguing in bad faith - perhaps the 5% is in favour of the destruction of Israel or pro a one state solution. What Bernie is arguing for is the end to the recent war which is popular with democrats and young people.
I get the feeling you're not really responding in good faith. Obviously some positions are far outside the Overton Window and it's pretty unreasonable for you to ask me to name some specific number of votes that a position should give before a party offers it as an option. "Implement/compel a ceasefire" for example is not outside the bounds of acceptability as you made out. Second, saying that Dems should, say, campaign on decreasing military aid to Israel is not saying that Dems should also campaign on giving that aid to Hamas instead.
I'll admit I'm somewhat incredulous about the support for Hamas that we see. From the perspective of Democrats, though, it seems electorally insane to offer any support for Hamas, and only qualified support for the Gaza Palestinians (who apparently like and/or support Hamas). You offered the perspective that it would be "smart" for Democrats to court that vote, and I pointed out two issues - 1) Support for Hamas is not popular in the US, even among Democrats, and 2) Support for Israel is far more popular, including among Democrats.
It's a losing proposition. I also don't think "chase single issue voters even if their goals are unpopular" makes any sense generally. It's perfectly fine for neither major party nor any minor party to support the KKK. It's also more than okay if no party supports Hamas.
I've never argued against someone pro-Israel on the internet before this, but I finally understand why the other side gets annoyed by online pro-Israel commenters strawmanning them constantly to the point that it honestly looks like the Israeli government is actually paying people (you) to do it.
There are a sizeable number of people who'd like their governments take actions that Israel, or at least the current government, might not like but are not pro Hamas.
Here is my peace plan for Israel/Palestine. It may take a few years, but it is still more realistic that any proposed alternative I have seen:
Every Jewish family should be legally required to have at least 10 kids. When those kids get 18 years old, the entire territory of Israel and Palestine becomes one huge democratic state, and everyone who lives there will get full citizen rights.
This is a perfect win/win solution for everyone. The Jews get their Jewish state. The Arabs get to live in peace at the place where their ancestors lived. Everyone gets democracy and human rights.
(Alternatively, make it 20 mandatory kids for every Jewish family, and also conquer Lebanon. You are probably going to do that anyway, so at least let something good come out of it.)
Sure, but for some reason it seems most people aren't really going for any of the many third way options. From my perspective, it seems the anti-Israel/pro-Palestinian side wants to force the issue to be binary in order to juice their gains. Mixed would by far be the most popular, followed by pro-Israel, if all options were on the table. The only way to get more than a lizardman level of support for Hamas is to force the dual choice.
I'm at the Whipple museum of history and philosophy of science at Cambridge university. So cool. Here's info about an instrument built by John Harrison to measure longitude at sea.
I've been to a similar museum in Harvard, but this is even smaller and less overwhelming...
Btw, Longitude is a great book for young kids.
Why does a kid even need to attend school if they can go to such a museum with guidance?! Only for a social life I suppose. And in some states because it's a legal requirement.
Great question. No, a museum won't be all of the curriculum, but a learning aid. And dies it have to be in your city? Give the $17k per kid per year to the homeschooling parents to use for such expenses.
The govt needs to be good at some minimal checks on the process and results. The results seem to be better than what public schools can produce. So this should be possible for motivated parents.
Different kids need different things, and the same kid might need different things each year.
I'm not claiming this is a perfect system. And others have thought through this better than I have.
Sort of but not really. The museum is chosen by the parents as part of a curriculum, one designed or children by the parents.
The teacher can be the parent (if they have knowledge in the area) or a homeschool teacher hired by the parent. Parents of homeschooled children with different areas of expertise form co-ops and teach each other's children based on their expertise.
The basic difference is if it isn't going well for the child, the teacher can be fired immediately.
I know this is probably very worrying for Americans because this is going to be your government, but today's headlines about Trump's cabinet picks have me laughing.
New Secretary for Education is, if I go by the headlines, "WWE wrestling mogul" Linda McMahon. I know there was a lot of disdain for his last pick there in the first administration, Betsy de Vos, so I'm awaiting (with a certain amount of glee, I have to say) the reaction to this. It would seem her experience in education was serving on the Connecticut state board of education 2009-2010 and being on the board of trustees for a Catholic school:
"She served on the Connecticut Board of Education for a year starting in 2009 and has spent years on the board of trustees for Sacred Heart University in Connecticut. Seen as a relative unknown in education circles, she has expressed support for charter schools and school choice."
More Catholics in the cabinet? I am delighted to see the Sinister Papist Conspiracy to extend the tentacles of the Vatican into American government is proceeding apace!
Dr. Oz to oversee health insurance? Okay, even I have to blink at that one.
Head of Commerce department could affect us here in Ireland, as (and I have to agree on this one), allegedly he tweeted (or Xed, is that how we should say it now?):
"Lutnick was recently critical of the nature of Ireland's trade relationship with the US.
"It's nonsense that Ireland of all places runs a trade surplus at our expense. We don’t make anything here any more – even great American cars are made in Mexico. When we end this nonsense, America will be a truly great country again. You’ll be shocked," he wrote in a post on X in October."
It is nonsense, and it's not anything tangible (except for the pharma plants here), it's all down to our tax strategy to attract multinationals (like the FAANG set) which have headquarters here and move money around from one bank account to another. The effect on our economy is such that we need to have a separate calculation for GDP to account for the 'phantom GDP' from this.
So I wouldn't be surprised if the Trump administration changes taxes to make it attractive for American multinationals to keep their business at home, and of course this will have a knock-on affect on Irish employment as/if plants are closed or the Silicon Docks crowd shutter their offices. We'll have to see what happens!
I'm curious: what's your view of the worst case scenario here? Like what's the worst possible outcome that a bad DoEd secretary could cause? In my view the DoEd doesn't do anything positive for the cause of education, so it's impossible to do any harm by nominating an incompetent secretary. I think the country would be better off if the entire dept was completely abolished.
The DoEd becomes dysfunctional without having its duties transferred to other agencies, six million college students (mostly from poor or working-class families) lose their Pell Grants, many of them have to drop out of school, their lives are diminished, the nation's supply of skilled labor is diminished, and there's a massive electoral backlash of mostly Trump supporters looking for the anti-Trumpiest possible candidates in the next election.
If you were under the impression that the Department of Education was just a bunch of wokescolds who send out letters to people trying to run good schools and saying "you need to be more woke or else!", then no, that's not how any of this works.
> six million college students (mostly from poor or working-class families) lose their Pell Grants
And? They're still college students. They are affiliated with the left as far as the populace is concerned.
If the economy does go to shit, they have more than enough scapegoats to blame it on. Hell, if left wing people end up leaving in droves, the administration won't even have to rig the next election in order to maintain power!
And why is that a bad outcome? In my view far fewer people should be going to college, *particularly* those from poor and working-class families. Elitism is good in elite institutions. Eliminating Pell grants would go some way to eliminating cost disease and bloat in higher education. Academia has become a lumbering zombie and a propaganda indoctrination zone. Anything that cuts it down to size is good in my view. Education needs far less regulation.
< In my view far fewer people should be going to college, *particularly* those from poor and working-class families. Elitism is good in elite institutions.
Well, Wanda, I come from a family with only one college-educated parent, and our finances were very tight indeed when I went to college. My parents were especially honest people, and did not fudge the truth to make us look poorer than we were when they filled out the financial aid forms. Our income and savings were low enough that I got almost a full scholarship for the entire 4 years. I graduated from an Ivy, scored 99th percentile on the GREs and got PhD. Jeez, maybe if you'd had Betsy DeVos's job I'd never have gone to college, and would instead be your hairdresser. Or, like, the lady who gives you Brazilian waxes.
As for other teens whose Pells you'd like to yank: I just looked at some data about four year colleges and their average SAT score.
For about 35%, the average SAT score is 1200 or above. 1200 places someone at 64th percentile in the group that took the SAT, & that group's average intelligence and academic performance is surely above the average for college students as a whole. I'd say it's someone scoring that well or better on the SAT is not in the in your born-to-be-your-hairdresser (or hair-plucker) group, and can certainly benefit from college-level courses. If you nix the Pell grants of all the kids in those schools, that's a lot of people unable to pay for an undergrad degree they could probably have made good use of. And only 8% of colleges have have SAT scores 1000 (the average score) or below.
Ok, you're the exception. Getting 99th percentile on your GRE's means you have a >120 IQ. You're exactly the kind of poor person that's supposed to be discovered via standardized testing.
>For about 35%, the average SAT score is 1200 or above.
Ok. An SAT of 1200 is an IQ of about 120 which is about the top 10% of the population. That's who should be going to college, not the ~50% that we currently send. I would MUCH rather sacrifice the 35% who should be there for the 65% who shouldn't. And pulling Pell Grants wouldn't prevent them from going, it would just make it more expensive. I want poor, potentially-stupid people to think very carefully about whether their mongrel genes can actually benefit from an expensive education before they just blithely go. College used to be a sorting mechanism. It's not anymore and that loss imposes very reals costs on society. It's incredibly useful for everyone to know where to get smart people.
>And only 8% of colleges have have SAT scores 1000 (the average score) or below.
Ok. That number should be 0. If someone gets a 1000 on the SAT then they shouldn't come anywhere near a college. College is for the top 10-25% of society. Everyone else should just learn a trade. Hairdresser is an excellent example.
<And pulling Pell Grants wouldn't prevent them from going, it would just make it more expensive.
"Let them eat cake."
Actually, the way the colleges set financial aid is by using the data the parents send about their savings and income, and the data parents must give is *very* detailed. There really is no loophole that allows people to look less able to pay for college than they are by having a secret rental property or whatnot. Of course they can lie, but colleges are allowed to do some checking around so that's pretty daunting. Using the data the parents sent, colleges calculate how much parents should be able to pay, and their idea of what people of a given income can cough up is generally more than people themselves feel able to cough up (though most can by penny-pinching. To cover the amount parents will not be expected to pay, colleges give a financial aid package, which is a mix of scholarship (i.e. college eats the cost), loans the parents or kid must take out, and money the kid must earn with an on-campus job. Financial aid packages always set the amount the parents and kids borrow as the maximum in government student loans they are allowed to borrow . Colleges give scholarships to cover the rest, and for kids who qualify for Pells the scholarship includes the Pell grants. So if the Pells are yanked, either the colleges would have to be willing to eat more of the cost or the kid and parents would have to borrow more. However, the family is generally already stuck with borrowing the maximum they are allowed to as federal student loans. The alternative for borrowing more is bank loans. People who are poor enough to qualify for Pell grants, and have already committed to take out $10,000 in student loans for the coming year generally do not qualify for bank loans.
<I want poor, potentially-stupid people to think very carefully about whether their mongrel genes can actually benefit from an expensive education before they just blithely go.
Jeez, Wanda, what even are mongrel genes?
< I would MUCH rather sacrifice the 35% who should be there for the 65% who shouldn't.
You know, I think I agree that people scoring at or below average on the SAT generally can't benefit from college courses. (Although there are college majors in things like hospitality, and when I was on vacation in Mexico I met a young couple who had actually majored in that and were looking like they were going to make a go of it with an eco-resort.). But I find you extraordinarily mean-spirited, and would even if I were not in the 35% you'd rather have doing your hair. It's pleasant to remind myself, and you, that nobody but you gives a shit who you are willing to sacrifice.
I'd prefer the backlash be more "reversion to the mean" and less "pendulum swings back like a wrecking ball", and I'd prefer not to have millions of innocent young men and women thrown abruptly out of school to that end.
And, yeah, among the hip rationalists and contrarians, academia is all cost disease and bloat and indoctrination and signaling. To the *vast* majority of Americans, the sort of social mobility represented by someone from a working-class family going to college and hopefully making a better life for themselves, is seen as a *good* thing. That's the story that made J.D. Vance a household name, but I suppose as an Ivy-league alum we're supposed to write him off as an irredeemably woke leftie or something.
The now-predominately-Republican working class likes that sort of social mobility, because those are *their* kids with a shot at what they see as a better life. The Democrats like that sort of thing because they think they'll get more Democrats out of it. I like that sort of thing because I'm a libertarian who believes in meritocracy. I'm not sure what you are that you don't see this as a good thing, but whatever it is, there aren't enough of you to matter.
And the bit where there's at least a large minority that's annoyed by all the kids going to Oberlin to study Uselessology or Columbia to join Team Hamas, fine, but *these aren't those kids*. Those are mostly the ones from upper-middle-class families who assume they'll be getting an upper-middle-class lifestyle just for having a college degree, and feel guilty about it. The ones who need a Pell Grant to make it through school, know they need to make that education count for something in the real world. So, yeah, pat yourself on the back for having reduced the number of college students, but you're getting rid of the students and your pissing off their parents and they're going to make you pay for that.
I think the problem is that university is associated in people's minds with being above average on the social ladder, so the popular sentiment is "if we all can get university education, then we will all be above the average".
Unfortunately, this is not how averages work. If everyone gets university education, then you will need PhD to get a hairdresser license.
So perhaps instead of extending the sweet deal of "get more credentials in return for a lifetime of debt" to an ever increasing part of population, we should go in the opposite direction and reduce the need of credentials. To become a hairdresser, you should need one month of training, at most. Make it so that if you decide that being an unemployed bum or a software developer sucks and you would prefer to be a plumber instead, you will only need two or three months, and you are legally free to start your own business.
Separately from that, we should improve education, and there are many things that could be improved there. But we should not make it legally required in order to get a good job.
What does "if we all can get a university education..." have to do with the policies of the second Trump administration, or with this discussion? Nobody here is proposing to give everybody a university education. There is no plausible outcome where everybody gets a university education in the next four years.
*Some* people will get university educations. Some people *should* get university educations. Not so much the people who are planning to become hairdressers, but definitely the ones who are planning to become engineers, doctors, lawyers, or businessmen, Ms (or Dr?) Tinasky may be of the opinion that those should all come from the middle to upper middle class and the proles should stick to their prole jobs like hairdressing. I'm strongly of the opinion that a good number of our future engineers, doctors, lawyers, and businessmen should come from the ranks of the working class and even some of them from the poor. And I'm pretty sure the wast majority of American voters agree with me on that.
Right now, the way that happens is with loans and grants that come through the Department of Education. I can imagine worlds where it doesn't work that way. Worlds where wealthy philanthropists or large charities or state governments handle that, and worlds where college is cheap enough that an ambitious young man or woman can pay for it with a part-time job. But we don't live in one of those worlds, and we're not going to build one of those worlds in the next four years.
So, if you break the Department of Education, then A: millions of students who I *think* you and I agree really *should* be going to college, will have to drop out, and B: a large majority of Americans will be really pissed at whoever made that happen.
If you've got ideas for substantive educational reform, great, but they're going to take more than four years. And to make them work as well as you hope, you're probably going to want to have the DoEd in good working order to either carry out the reforms or to seamlessly transfer operations to whoever is going to carry out the reforms.
>To the *vast* majority of Americans, the sort of social mobility represented by someone from a working-class family going to college and hopefully making a better life for themselves, is seen as a *good* thing.
Those people are wrong and I don't think that the federal government should be in the business of laundering their delusions. The "everyone should go to college" idea is just a hedonic-treadmill effect. The government shouldn't fund that.
If you're a working class kid who has class-bucking IQ then fine, suck it up and take on some debt. College isn't the express-lane to upper middle class that it used to be, so low-class families should legitimately stop and think about whether it's the right choice. Eliminating federal grants is something that would encourage a more rational cost-benefit analysis among the consumers of higher education. College was designed to be for the top 20% or so of the population. Anything more than that is just a waste of resources.
<If you're a working class kid who has class-bucking IQ then fine, suck it up and take on some debt.
You seem to be very underinformed about paying for college. Did you think Pell grants cover everything? Nope, they average about $5000 per year. And the average kid with a Pell grant ends up with a 15% *larger* debt, approx $26,000 total, than kids who did not qualify for a Pells, presumably because the Pell does not fully make up the difference between their funding and other kids'. So be of good cheer, Wanda, the Pell kids are doing plenty of sucking it up.
I can't tell what Peter is trying to say, but: people do want social mobility, of the type that comes from being capable of doing things that pay well enough to fund their entry into higher social strata. What they *don't* want is the type of social mobility that only comes about by bending the knee to the people currently in the top stratum because they have arbitrarily seized that stratum and have changed the entry criteria away from being capable of producing actually valuable things. Possibly taking Pell grants hostage in the process.
> And why is that a bad outcome? In my view far fewer people should be going to college, *particularly* those from poor and working-class families.
Definitely because nobody wants social mobility. The real problem is the rich bit dim, and what is needed is a reduction in certain courses but that’s less likely to happen.
Seems like Trump is in an awkward place. He put a number of normal people with regular credentials in his cabinet last time, and not only did he get lots of flack about it anyway, but lots of the people he picked tried to foil him. So he's highly incentivized to bypass all the normal picks.
I don't know what people expected differently, to be honest.
"Huh, everyone who is reputed to know what they're doing thinks I'm a moron and tries to derail my plans. But since I'm obviously infallible, I guess I should go with people who everyone says are clueless next time". Peak Dunning-Kruger right there.
Maybe, or maybe there really is a "deep state" of people with similar views across party lines and they hate Trump for his opposition to their one-party rule.
I'm not saying that's necessarily true, but I also don't think we can rule that out. I'm troubled by how many things Trump got criticized about that Obama did before him and Biden did after. Kids in cages being an obvious and well known case, but also Biden continuing Trump's tariffs and things like drone strikes against enemy leaders.
Let me rewrite the first sentence from a different perspective: Maybe there really is a group of public servants with a deep knowledge of how the machine runs and commitment to keep the state running regardless of party affiliations, who really hate Trump for his blatant stupidity and disregard for reality and morality.
Of course, that's the optimistic spin, and the reality may be somewhere in the middle. But I really believe that any complex organization needs a lot of people with specialized knowledge, and replacing them will lead to dysfunction. The question is, "is the current state of things so rotten that it's worth tearing big parts of the system down and having effectively nothing in their place for a while?" Apparently, in the opinion of Trump, Musk and others, it's worth it. My guess is that they're going to replace somewhat corrupt but functioning systems with ones that are both dysfunctional and utterly corrupt... but we'll see.
That is a really good question, and I also do not have the answer to it.
I do have a problem with leadership in a democracy ignoring the well-intentioned pleas of the population. This includes people on the left (Bernie supporters) as well as people on the right.
Many of the problems are not exactly subtle or hidden. That outsourcing and deindustrialization has gutted the entire rust belt is not news to anyone. That the leadership class seems to think of this as somewhere between fine and good is a serious problem. What's an economy for? They seem to think that number-go-up is enough. Serving the needs of the population is a better answer. If number-go-up also helps the population (as it did in the 50s) then everyone can be satisfied. If instead it helps people in certain fields to a massive extent but most people are hurt or gain little, that's going to cause problems.
I don't think Trump is a cause, but a symptom. I've been listening to working class Republicans hating on their elected Republicans for well over 20 years. It was only a matter of time before someone showed up to do what those people actually wanted, instead of more wars and more international trade. If the solution ends up being bad, which is entirely possible, then that still doesn't fix the underlying issues. The only thing that will fix the issues is those so-called experts to take the needs of the lower classes seriously.
I assume that those working class Republicans you talk about hate on their elected Republicans in between elections, but vote Republican at election time.
The Biden Administration passed the CHIPS act and the Inflation Reduction Act, and we saw the election results. It’s possible that the policy would have been a political winner with better messaging; I personally don’t have any expertise in political messaging so it’s hard for me to say. It’s also possible that the policy helped politically, but not enough to overcome other factors which allowed Trump to win. But it suggests that addressing deindustrialization is not a politically winning policy.
Indeed, you go on to write that these Republicans voted against more wars and more international trade. The first has nothing to do with economics, and the second is different from the issue of outsourcing and deindustrialization gutted the rust belt. If you put tariffs on the import of raw materials, that makes American manufacturers less competitive, so it reduces international trade both by reducing the import of raw materials and by reducing the export of manufactured goods. Great if your goal is to reduce international trade, bad if you care about American manufacturing.
Based on polling done in 2019, Bob Altemeyer wrote, “The biggest reason by far that people supported Trump was their level of prejudice, but some (not many) relatively unprejudiced subjects approved of him because they thought they were prospering thanks to him.”
I see what you mean, but I am not sure the wishes of the electorate can be fulfilled in a straightforward way without unintended consequences hitting them hard years or decades later. My understanding is that the bleeding-out of parts of America's industry was an inevitable side effect of globalization (i.e., free trade and Pax Americana), which not only lifted hundreds of millions of people worldwide out of poverty, but also brought about relative peace, a ridiculous pace of technological progress and, last but not least, the possibility of bying cheap stuff from China. Rolling it back may or may not lead to workers in the US having plenty to do (depending on how it's handled), but it very likely leads to an overall lower standard of living in the US and probably, within a few years, lots of wars leading to a thorough disruption of global trade, neocolonialism, biblical-level famines and general unpleasantness which eventually impact everyone, including the workers in the US. (Peter Zeihan paints the full picture in "The end of the world is just the beginning"). People who have worked for decades in the State Department probably understand this; the guy whose economical platform consists of "I LOVE tariffs" maybe doesn't, and neither do many of the voters who can't find China on a map.
Curious which picks of his you think got a lot of flack while being normal. At least some of his picks were quite popular - James Mattis for SecDef was confirmed 98-1 and H.R. McMaster for National Security Adviser was confirmed 86-10, for example.
He got burned in two different ways. Careerists who sailed through confirmation tended to stymie his plans. Anyone else got pilloried as unfit even before they got in the job.
Would we really expect Biden, Harris, Obama, or Bush to keep playing by the informal rules when those were the results?
I mean, if your choice is between "qualified people who will easily be confirmed but won't do whatever I tell them to" and "widely-criticized people who will do anything I want", I think MOST people would wonder whether the first group actually are doing a better job than the second group. There's more to cabinet roles than rubber-stamping the President's agenda, and surrounding yourself with a bunch of yes-men could be pretty self-defeating.
Sure, but Trump ran on a platform of doing things differently than other politicians. His options are to either follow the consensus like every other president and thereby betray his voters, or put people in his administration that aren't part of the consensus.
I think there's a lot of room for us to discuss which examples are "non-consensus" and which are "unqualified cranks".
E.g. Rex Tillerson for State was definitely not a direction I would have gone in, and I would argue was pretty meh for the department, but at least was an international businessman and probably falls into the "not willing to do whatever Trump wants" category. RFK, Jr. is a lunatic conspiracy theorist with the opposite of qualifications who could do astonishing damage to public health.
Yes, I think this is a factor in Trump's more controversial picks for his next cabinet. It's pretty clear that he felt burned by some of the "normal people with regular credentials" in his first administration, so this time he's putting a premium on reliable allies and friends. Trump has been close to the McMahons for decades, so no surprise he's turning to Linda again, this time in a larger role.
Conservatives have for years wanted to dispense with the federal Department of Education; it's not immediately clear whether Linda McMahon is being put in charge to try to carry that out. Certainly the Project2025 guys want it to happen as described in their big blueprint document.
The McMahons, husband and wife, are actually a helluva shared life story which some competent writer could make a compelling book out of some day. A couple since high school, early in their married life they were on food stamps and a bit later went bankrupt; neither had been born into any money. Despite which they became very successful and will soon celebrate their 60th (!) wedding anniversary. Frankly they are in some ways what Trump cosplays as....Linda held a lesser appointment during the first Trump administration, and once won a party nomination for a Senate seat. So she's not now a complete rookie to the political/governmental big leagues.
Lutnick is someone who's pursued an admirable adult path after a rough start (his mother died of cancer during his last year of high school and then his father was killed by medical malpractice during his first week of college). His qualifications for Secretary of Commerce are not obvious but then again that job has become largely toothless anyway; his appointment isn't much different than our longstanding shitty practice of naming large campaign donors to be our ambassadors.
Meanwhile though I am not at all laughing at Trump naming his literal personal defense attorneys to the #2, #3 and #4 positions of our Justice Department. _That_ is a new one which not even the likes of Tricky Dick or Slick Willy or Harding's handlers ever tried.
McMahon converted from Protestantism to Catholicism. Vance is also a Catholic convert. You'd think Protestants would be bothered by losing so many of their future elites. Or maybe they're proud of being a religion for low-class people.
Mass immigration from third world means internal white difference are no longer relevant. No one cares if you are an Irish (American) or Swedish (American) or just a good old Anglo. When a third your university class consists of recent arrivals from China and India, another third children of Latin American immigrants, you are just "white". The massive secularization starting in 2000s means the same process has happened with religion.
> So I wouldn't be surprised if the Trump administration changes taxes to make it attractive for American multinationals to keep their business at home
I thought the point of Ireland was that it was the cheapest (tax-wise) European state to be in? If I recall correctly, there is a significant tax advantage (or requirement?) for big tech companies to keep a European bank account if they’re doing business in the EU…
That's what Trump's guy is complaining about, and others before him. The EU told Apple to pay us €14 billion in back taxes, and our government had to be *forced* to take the money, they tried putting it off because that's our one advantage, our low tax regime for foreign investment.
As I get older it seems that cats, in general, like me more. When I was a teenager, even friends' cats would give me a wide berth. Then progressively catkind got more neutral about my presence. Now it's a semi-regular occurence that cats actually come up to me and ask to be petted. Even completely strange cats, in the street. I don't think my behavior toward them has changed (I'm not aggressive but reserved. Since I'm mildly allergic, I avoid touching them). Is it because they see fidgeting as scary and I got more immobile with age? Has anyone else noticed it happens to them too?
It seems to be cats specifically and not pets in general.
A primate grin is NOT welcoming to a cat, quite the reverse as one is showing teeth. Most cats are also averse to wide open looking eyes, especially staring at them, as they view these as a challenge. Also, outstretched human arms tend to alarm them, especially when these move abruptly.
A cat smile is a slow blink and gently narrowing eyes. So maybe your eyes have become apparently less open and alert looking as you age, possibly behind glasses you never wore as a teenager, when cats were avoiding you. Either that, or you regularly eat toast or muffins for breakfast. Cats can overlook most worrying quirks of human anatomy if the humans smell slightly of butter! :-)
You may have cracked it, I have a drooping eyelid and now that you mention it, in all three of the last instances the cat approached me from that side. Never saw (heh) the pattern.
Why isn't it a bigger deal here that the benefits of scaling AI larger seems to have hit the ceiling? Shouldn't that be the biggest AI story in the world? Why aren't chip and utilities stocks crashing?
I know they say blah blah blah we got other ways to improve AI --- but fucking shit --- the way to improve it so far has been scaling on bigger data sets. Why should believe another way of scaling is going to improve things all of a sudden?
Nine months ago people around here were going "Why isn't the market reacting to the fact that AI will soon be godlike and/or destroy us all?" But people outside this bubble never really expected that to happen.
Because the NVIDIA etc stock bubble hasn't burst yet.
Stipulating for the moment that AI has hit something ceiling-ish: Hype sells. There's stories to be told about hype, and people want to tell, hear, and be part of those stories. Blood also sells, and that includes the green-tinted financial sort of blood. The interregnum where the hype has worn thin, but the last round of Bigger Fools haven't given up and sold out yet, that's just kind of meh and boring.
And around here, we've basically got people who believe AI is wonderful and awesome, and we've got people who believe AI is an awesome and terrible danger, and we've got people who think the first two groups have been talking about AI way too much. If it turns out that AI stagnates at "meh", then for both the wonderful-and-awesome and the awesome-and-terrible groups, talking about AI would mean talking about how they were wrong. Nobody wants to do that. And the third group has had all the AI-talk they care for so they're not contributing more.
Sooner or later, either AI will break through the "ceiling", or the bubble will burst, and either one of those will be a big deal here. In the meantime, all is quiet on the AI front. Enjoy it while it lasts.
>Sooner or later, either AI will break through the "ceiling", or the bubble will burst
A third, "meh"er possibility is that LLMs stagnate - but the current level winds up being good enough for a lot of applications (which may include enshitification of customer service). Remember how long it took for businesses to work out just how to use personal computers intelligently!
>if AI paused tomorrow, people would be figuring out applications for decades.
I do think we can rule out a near-term stagnation of machine learning on _trustworthy_ data, e.g. the protein fold prediction. That doesn't have the hallucination problem that predict-the-next-token has.
Yes, this has all the vibes of 1999 internet bubble. Both things will likely come to pass: the bubble will burst, and a useful machine-learning infrastructure will be available to used for many more years.
Yes, that sounds plausible. And, when the internet bubble burst, the stocks came down, but the internet didn't go away, as if the whole technology was a bad idea, so I agree with your expectation that machine learning applications will persist.
I'm pretty sure LLMs stagnating, even at a useful level, still results in a spectacularly newsworthy bursting of a bubble that's currently valued on the expectations of LLMs very much not stagnating.
I'm kind of surprised that the people most concerned with AI singularity haven't attacked it more from an energy conservation perspective. The amount of energy required to keep improving AI is truly immense, and it shouldn't be hard to make a basic environmentalist argument against using that much energy on AI, even putting aside concerns with possible AI superintelligence. I'm surprised that more people in *this* social circle haven't done so, at least from what I've read so far.
Probably most people here are already concerned with anthropogenic climate change. And much of that climate change comes down to energy consumption. I have some very leftist friends that are highly anti-crypto currency precisely because of how energy-intensive crypto mining is (at least from what I've heard). Well, it sounds like AI is becoming even more energy-demanding than crypto-mining is.
If you're concerned about anthropogenic climate change, and if you're concerned about AI superintelligences, then it seems there's now a great way to tie them together and attack both at the same time. It seems odd to me that I haven't read this argument put forward more.
As I understand it, OpenAI's GPT was developed essentially by simply ploughing through every recorded utterance since the dawn of time through 2018 or whenever it was. On that assumption, If energy is a limiting factor then presumably further progress can be made only by structuring the data better, into a network of more self-contained units, and enhancing these independently.
If we draw a rough parallel with programming, the first operating systems were presumably masses of messy machine-specific assembler, but over the decades these have become far more modularized at the same time sophisticated.
Well, when the same people are pushing for electric cars, it looks transparently dishonest to be demonizing electricity consumption in non-Woke applications.
Pushing for electric cars happens when the choice is between many little engines burning fossil fuels to accomplish some task and using centrally generated electricity to do it.
When instead the choice is between doing a task and not doing it at all, yes, the same people will push for the latter option.
No-one's arguing for clockwork+petrol AI; Nvidia is not, last I heard, building giant Babbage engines for running LLMs, though I do admit that would be kind of awesome; and no-one's putting petrol engines in their data centres.
Meanwhile, the people pushing for electric cars do also very much tend to push to reduce car use overall.
All seems consistent to me. Where's the dishonesty?
In pretending that the electric cars won't require electricity generation and transmission infrastructure enhancements along the same lines as for AI or any other non-Woke-favored use.
>Meanwhile, the people pushing for electric cars do also very much tend to push to reduce car use overall.
That _is_ a fair position, and characterization of their position. I generally oppose it, but it is at least honest.
> I generally oppose it, but it is at least honest.
I do feel there is a massive disconnect between people who live in very densely populated areas where owning and using a car is misery and people living in suburban sprawl and/or rural areas, where a dense public transport network would not see enough use to be feasible and so there is no sane alternative to car ownership; people in each of the two groups who have not lived in the other kind of place find it very hard to get in each other's headspace.
> In pretending that the electric cars won't require electricity generation
Does anyone pretend that? I've only ever heard it as a MAGA talking point, trotted out while ignoring whatever the interlocutor actually says; much as you just did to me.
Many Thanks! I've usually seen it as the _absence_ of discussion of all the electricity generation and transmission requirements in some electric car advocacy piece. Look, I've seen Greens block _transmission lines_ , the national equivalent of an extension cord, let alone power plants. Now, these may not be the exact same people who are pushing electric cars, but, as an overall political group, in aggregate, they look hypocritical.
Frankly, for the training phase part of AI development, solar or wind could potentially be used. Training is an incremental process. Nothing intrinsically horrible happens if the process runs during daylight hours. There is an economic trade-off due to having the capital cost of the chips lying fallow at night - but if we _really_ are in or near a regime where energy costs dominate, this should be ok. Solar _without_ storage is one of the cheapest, possible _the_ cheapest, source of electricity.
If we _aren't_ in a regime where energy costs dominate, then these energy discussions look like a dishonest cudgel being used to bash AI which wokesters actually oppose for _other_ reasons.
I don't think most people in AI -- myself included -- think we have hit a ceiling. Everyone in the world is compute capped. We need more chips, and more energy. But chips and energy are continuously getting better and cheaper. It shouldn't be lost on anyone that tech basically lobbied the government to make nuclear a green energy category and starting buying a bunch of micro reactor time as a result. That should give you a sense of what the industry thinks of as our current limitations.
Related: there are several good papers on scaling laws that show that, following a power law, exponential improvements in compute and data together result in linear improvements to models. Anecdotally, however, linear improvements in models lead to exponential improvements in acquired abilities and outcomes. This is part of why everyone is compute capped. Every single GPU is already going towards training the largest models possible, they need more to scale more.
(Inb4: Data is not the limitation -- Google has tons of data they are purposely not using, no one has the compute capacity to really use images and definitely not video or audio.)
"Employees who tested the new model, code-named Orion, reportedly found that even though its performance exceeds OpenAI’s existing models, there was less improvement than they’d seen in the jump from GPT-3 to GPT-4.
If this were just a few hedged anonymous reports about “less improvement,” I honestly wouldn’t give it too much credence. But traditional funders and boosters like Marc Andreessen are also saying the models are reaching a “ceiling,” and now one of the great proponents of the scaling hypothesis (the idea that AI capabilities scale with how big they are and the amount of data they’re fed) is agreeing. Ilya Sutskever was always the quiet scientific brains behind OpenAI, not Sam Altman, so what he recently told Reuters should be given significant weight:
Ilya Sutskever, co-founder of AI labs Safe Superintelligence (SSI) and OpenAI, told Reuters recently that results from scaling up pre-training—the phase of training an AI model that uses a vast amount of unlabeled data to understand language patterns and structures—have plateaued."
What to make then, of the claim that researchers have unlocked a new axis of scaling by giving the models time to “think”? For this is what the entire industry will now rely on, both in practice but also to keep the sky-high hype alive.
I think this new axis of scaling will matter far less than people expect and will not keep the train of improvements chugging along anywhere near the previous rate.
First, this new type of scaling appears limited to certain domains: it makes the models no better on English language tasks or provides only minimal improvements for questions about biology, for instance. Second, there are good reasons to believe that this sort of scaling will be much more expensive: linear improvements for extreme sums of money."
As tech journalist Garrison Lovely wrote about it on Substack:
If you have a way of outcompeting human experts on STEM tasks, but it costs $1B to run on a days worth of tasks, you can't get to a capabilities explosion, which is the main thing that makes the idea of artificial general intelligence (AGI) so compelling to many people…. the y-axis is not on a log scale, while the x-axis is, meaning that cost increases exponentially for linear returns to performance (i.e. you get diminishing marginal returns to ‘thinking’ longer on a task). This reminds me of quantum computers or fusion reactors—we can build them, but the economics are far from working.
But I think people focusing on price or the domain-specificity of improvements are missing the even bigger picture about this new supposed scaling law. For what I’m noticing is that the field of AI research appears to be reverting to what the mostly-stuck AI of the 70s, 80s, and 90s relied on: search.
Saturation along a straightforward axis is indeed always bad news.
I'm not as pessimistic about the combination of LLMs + search as Hoel is.
LLMs have always seemed to me to be like system 1 thinking - pattern recognition on steroids. And search is similar to system 2 thinking, articulating possibilities, then evaluating their relative merits.
I'm _not_ a worker in this field, so my gut reaction is ill-informed, but these _look_ complementary, and might still be a pathway to AGI.
I do think we can reasonably say that _some_ sort of artificial neural net must be able to reach AGI, since _we_ are neural nets, and _we_ exhibit general intelligence. But, this doesn't say that the _current_ artificial neural net interconnection _architecture_ can reach AGI, and exploring multiple alternative candidate architectures, at many millions of dollars of training costs per experiment, might be a deal-breaker :-(
I'm not one to go against Sutskever, his track record is certainly better than most. That said, I'd also be curious what, exactly, he's doing over at SSI, and how much he stands by how Erik is interpreting his words.
More broadly, 10x-ing the amount of compute and data that is fed into a GPT is no small task and I'd be somewhat surprised if anyone has gotten there. Is Orion 10x bigger than GPT4? The latter is almost 2 trillion parameters trained on 25000 GPUs using 13 trillion tokens. Is Orion 20 trillion parameters trained on 250000 GPUs using 130 trillion tokens? I'm really doubtful.
But such a model may well be linearly better than GPT4, in the same way GPT4 was linearly better than GPT3. There's just massive engineering challenges that need to be addressed first. I don't think Sutskever can solve those engineering problems with SSI, the company is not big enough. Note too that Sutskever was sorta probably pushed out of OAI, he didn't leave out of some scientific disagreement. Meanwhile, Microsoft (OAI) or Google might be able to actually hit the level of scale necessary for a real GPT5, and they are both certainly trying (again see the nuclear power expenditure). But getting your hands on 250000 GPUs (or some number of TPUs for Google) is tough! So imo it remains to be seen that the scaling hypothesis is truly dead. And my own circles of AI researcher friends sure seem to think there's more juice to squeeze.
There will always be people willing to throw money at fancy new tech. And the big companies have set out their stalls with AI versions of their own, from Microsoft to Google, all producing consumer versions to "try this and let it help you".
So there is too much money already spent on this to stop, or for share prices to fall. The faith is that AI progress will keep going, or even if it doesn't, that widespread adoption of it will keep the profits rolling. Internet of Things Mark II.
It's not that big of a deal, because even if the scaling laws do hold, they imply that you need exponentially increasing amounts of power and compute to eke out marginal gains in ability. What good is an AGI if you need one data center and one power plant to replace one human employee? Or even 1,000 humans?
I think it's pretty obvious that a pure transformer approach can't lead to true AGI, because we have yet to find a way to train models on small data sets, and there's only so much you can put in the context window. For a lot of tasks, there's just not enough data available to train an LLM on, you have to "learn on the job", and a pure transformer can't do this.
OpenAI's o1 approach is much more promising and shows that scale alone isn't everything. I'm somewhat sure that LLMs will be part of the first AGI, however.
"we have yet to find a way to train models on small data sets"
Humans have GI (General Intelligence), and I think THEY don't actually work on small data sets. Humans take in an awful lot of data points every second. Just the eyes have about 120 million rods and cones, each of which are collecting data many times per second.
Scaling up a process is linear, no matter how you do it. Economies of scale can reduce per-unit costs, but linearly. LLMs are an important milestone, being able to do things computers weren't expected to be able to do, but it will take another kind of technological leap to have geometric growth in AI.
We are born wired to pay attention to language, and to use the language we learn to make sense of experiences. We are wired to experience it as carrying meaning. We have a special brain area set up to organize and remember the language we hear, and find regularities and remember them. We are born with a craving to make sounds with our mouths, and to imitate sounds we hear. we are born preset to transfer what we know to other modalities —
writing, sign language, foreign languages. All that gives us an enormous advantage over a machine that has to figure out stuff from scratch.
Agreed on everything except your last sentence. Yes, we have, e.g. Broca's area.
But the language processing is the part of AGI that is a _solved_ problem at this point. Probably inefficiently solved, but solved. As you showed with your experiment with the pendulum problem, they are still getting very simple physics (what bumps into what) wrong. And I keep seeing them get simple chemistry questions wrong. But these aren't language problems.
Yeah, you're right about my last sentence. In fact it's odd to be denigrating AI about language, when it seems like that's the thing where it comes closest to being our equal. When I talk with GPT4 now, I don't even think about simplifying, just write my questions or comments exactly as I would here, and it understands fine. What I had in mind, but did not say, is that our inborn ability to master language is set up so that learning language and learning about the world link up easily, in fact they are sort of two sides of the same coin. Here are a couple small cute examples. : When my daughter was learning to talk we had 2 affectionate cats who hung out with her a lot. One day she was touching something soft -- I can't remember what -- and she said "it's kitty." And she didn't mean it was a kitty, she meant it was soft, as cats are. Learning the textures of things, and learning words for them, were part of the same process. Not knowing the word *soft* she felt free to invent a word by extending the meaning of a word she knew, and knew I knew. Here's another example, sort of a cousin of the first, I think from when she was 3: We were in a Starbux and slow jazz was playing, and I asked her if she liked the music. "It's whiny," she said.
So the hard-to-duplicate thing about our species isn't that we are set up to learn language, and so can do it with way smaller samples. It's that we're set up to learn langauge *and integrate that* with learning about world. So the way we know what pendulums do comes both from things we've read or been told and from handling objects.
Many Thanks! Yes, the LLMs have a lot of trouble with things that we learned with our senses, and that we, as you said, _linked_ to language.
>It's that we're set up to learn langauge and _integrate that_ with learning about world.
Yup. And a good chunk of how humans learn is by acting on the world, both physically and socially, and seeing the reaction. As you said:
>When my daughter was learning to talk we had 2 affectionate cats who hung out with her a lot. One day she was _touching_ something soft -- I can't remember what -- and she said "it's kitty." And she didn't mean it was a kitty, she meant it was soft, as cats are.
[emphasis added, and also agree about the generalization she did]
Reading enormous amounts of text can substitute for _some_ of this, but LLMs clearly have problems in these areas, and it isn't clear how far into something like robots-who-can-play we would need to go to get around them.
I dispute this assumption. We are born with awesome pattern-matching brains, which recognize sounds, gestures, differences in pitch, and myriad other things, and determine ways to use some of these things to communicate with others. Language is learned behavior.
Machines, on the other hand, are nothing but Chinese rooms so far. Even the idea that they interact with language must be programmed into them. Any dataset can be interpreted any number of ways, and it is up to the programmers to choose which ways.
It’s well known that our brain is set up to attend to and learn language. Really. If you doubt me, google something like “is human brain optimized for language.”. iIf you still doubt it, read the links.
I don't think you understand me. Yes, the brains perceives the components of language. But "language" itself is for communication, which is something the brain determines for itself. We do know an awful lot about how the brain handles such things (https://en.wikipedia.org/wiki/Language_processing_in_the_brain).
Consider a thought experiment of taking some people as babies, and raising them without teaching them any known languages. They will likely develop their own languages to interact with each other, and making mouth sounds will probably be only one part of that. If a baby were abandoned in an isolated area and somehow survived, it would not develop any language capability, until it conceived the desire to communicate, perhaps with animals it befriends.
So I'm saying that the brain is CAPABLE of language, and is good at it, but it isn't really programmed in any way for language specifically. A bicycle is designed for travel. It could be placed into a device that converts the energy of motion into electricity. Even though it generates electricity well, it remains designed to convert a person's energy into wheel rotation.
> Humans take in an awful lot of data points every second. Just the eyes have about 120 million rods and cones, each of which are collecting data many times per second.
I don't think this argument works, because almost none of that sensor data is used for text-based learning. When a human reads the definition of a previously unknown word plus one example (corresponding to maybe 50 tokens), they're able to use it from now on. It doesn't matter that the eyes are collecting Megabytes-equivalent of data while reading that short text, because that sensor data still only contains about 50 tokens worth of training data.
An LLM is able to temporarily "learn" to use a new word this way, as long as its definition is still in the context window. However, to permanently learn this word (i.e., to modify the model itself), a lot more examples and training data are required. Maybe a sufficient amount of synthetic training data can be generated by the model itself from the definition in the context window; however, I don't consider this a "pure" transformer architecture anymore, but rather a system with the LLM as one of its components.
You're thinking of learning like a computer or data scientist. Living things don't think that way, and though it may be inefficient, it is certainly effective. LLMs analyze text because they have a format of data analysis programmed in that allows symbols to be correlated in that way. Brains, however, figure out HOW TO CORRELATE THE DATA, and which data to correlate.
In other words, which pieces of data should be included in analysis? We make the decision for LLMs. Imagine giving LLMs long strings of bits to analyze, say 2^100 bits long. What is it to make of it? We could have encoded a bunch of pictures in a certain format, or encyclopedias of informational text, or some format of blueprints of various buildings, or recordings of various sounds. A brain figures out what the sensory information means, but a LLM must somehow be "told".
That's essentially what the "attention" layers in the transformer architecture do. And the adjustable parameters in them get optimized during training, along with the other parameters, so, yes, LLMs do learn
>which pieces of data should be included in analysis
Well, it's a pretty good place to store all the Europeans so they don't come over here and be angry at us.
Also, there are still plenty of places in Europe where people don't generally get angry at Americans. Thought I'm certain they'd be willing to make an exception for certain Americans.
As other respondents said, I think the basic premise is mistaken (or very context-dependent at best). I can think of a few unserious answers as to the purpose of Europe.
1) Give Americans places for cheaper (yet chic) vacations where you can drink tap water.
2) Showcase how to properly be left-wing (think of how many meanings this sentence can have!).
3) Give the World Spirit some psychological drama for this grand story of stories that They’re realizing.
4) Give more data and customers to tech companies (there are more Facebook users in Europe than in the US).
5) School of life for the US “greatest generations”. Why don’t Americans cook up another big European war again? (As everyone knows*, the UK only got the atomic bomb because France did. You can still rekindle the 900-old rivalry!)
*Well, that’s what actors playing top-level UK bureaucrats in a comedic 1980’s TV show say in-character anyway. Why would they lie?
What makes you think they get angry with us when we go there? I've been there multiple times in the last 12 months, I didn't encounter anyone who seemed angry at my presence.
I think they mean there's a bunch of people in social media saying "argh tourism is bad", and anecdotally I've seen graffiti in more than one place I traveled to saying "refugees welcome, tourists go away" (granted in the rest of Europe, not in England).
I think this sentiment is just people blaming bad economic conditions on a scapegoat, and it's probably a minority of people. If tourism actually decreased many European cities would actually suffer.
I have read that high profile destinations like Venice are getting overwhelmed by large scale group tourism from countries like China whose citizens aren’t well socialized for tourism (yet). Also, social media drives large numbers of young people to take their selfies in the same Instagrammable locations, causing bottlenecks.
There's people on social media saying literally anything you can think of at any given moment. I'd save the worrying about people being mean to you when you vacation in Europe for when they are, actually, mean to you when you vacation in Europe.
Chances of random Euros being "angry at Americans when they go there" have never been particularly high as far as I know, but is probably lower now than during the Iraq War.
The UN decided to do a worldwide survey, with just a single question. The only question asked was: "Would you please give your honest opinion about solutions to the food shortage in the rest of the world?" The survey was a huge flop.
In Africa they didn't know what "food" meant. In Eastern Europe, they didn't know what "honest" meant. In Western Europe, they didn't know what "shortage" meant. In China, they didn't know what "opinion" meant. In the Middle East, they didn't know what "solution" meant. In South America, they didn't know what "please" meant. And in the USA, they didn't know what "the rest of the world" meant.
The software actually works. I have all sorts of negative feelings about Turbo Tax in specific and regulatory capture in general, but they found me a deduction once that will cover all the money I'll ever pay in Turbo Tax fees for the rest of my life.
It's not like Turbo Tax is comparable to Microsoft Office applications. Excel and Word have inertia because they are the standard we share with each other in the business world.
We don't need to share our Turbo Tax with each other, right? So why do people keep turning to it instead of a cheaper alternative?
They can prepopulate much of my data as a returning customer, and I can typically do my taxes each year in less than an hour. For me, it's not worth the search cost to find something cheaper to save $100.
Also, I've heard it alledged that TT lobbies for more complex tax laws to help maintain their confuseopoly position in the market, but not sure of this.
Regardless of your politics, what are people doing to prepare for the incoming administration? Buying crypto or real estate to get around inflation? Getting all your vaccinations done ahead of schedule in case access is disrupted? Buying organic food in case food safety is disrupted? Nothing in particular?
I bought oil and gas stock index and partially sold long-term (20+ years) US government bonds index. Also I've started to incrementally increase my S and P 500 allocation.
Epstein for Secretary of Education! Oh, he's dead? OK, that old girlfriend of his who's in jail, how about her? And I know it's a bit weird, but how about a dung beetle for Sec. of Health. And what about that guy who wore a Viking helmet in the Senate for Sec of State. Fearless, you know?
* Bought Series-I bonds up to the purchase limit for myself and my wife.
* Decided to take the saver's credit this year instead of realizing capital gains, as the Trump tax cuts aren't going anywhere for at least another four years.
To use them on who, yourself or others? Because if it is the latter, I highly suggest against it! You'll be completely fine if you don't do anything stupid. And you're not an illegal immigrant. Or a legal immigrant who's black. Anyways, even if that was the case, fighting back will just get you killed faster. Best to just stay put and watch the fireworks.
Ahahaha. Well, I hope you realize that the police and military usually outlast the economy. Their loyalty is the number 1 priority for any government, after all.
Any trustworthy analysis on why the final Selzer poll in Iowa (which showed Harris *ahead* by 3 percentage points) wound up so wrong (Trump won by 13 percentage points or so)? Selzer has a long and impressive track record, and it doesn't look like there was a dumb spreadsheet entry error or anything of that sort. Did the poll just get massively unlucky, on a level analogous to "we got p<.005, but the null hypothesis was in fact correct; this was the 5-in-a-thousand chance"?
One of the wackier conspiracy posts about the election I've seen is that Selzer's poll was correct *but* Musk and Trump hacked the voting machines and stole the votes, which is why the results of the poll didn't tally with the election results.
That's some religious level of faith in "Ann Selzer is always right" there.
On the other side of the wackiness gap, I've also seen suggestions (but how serious versus tongue-in-cheek I have no way of assessing) that Selzer was going to retire anyway, so she decided to burn all her credibility to get the result she wanted. Release a poll showing Harris with a massive lead, hopefully discourage all the grubby MAGA voters from turning up and energise the Brat Summer set to turn out and be part of the victory.
I don't believe either set of alternatives, but I do lean towards "Selzer was getting the results she might have subconsciously wanted" as one reason why the polling was so out of sync. Even pollsters are only human and have their biases.
The analyses I've read emphasize that her polling method was unusually "hands-off"; that is, she adjusts for very few factors. I found an old 538 article saying merely "age, sex, and congressional district". It turned out that, in this election, there was an effect that strongly confounded that uncontrolled analysis - I would suspect education as the culprit, given that the college/no-college split widened this year, and that education correlates with polling response rates. There was probably also some nontrivial amount of random error, though I suspect it didn't have to be on the order of thousands-to-one.
I think if there's an aberrant thing here, it's people being overimpressed by her track record - overcorrecting based on a small set of dramatic events, starting in 2008. No shade to her, really - she's always been transparent about her method, why she likes it, and its limitations - but it should have been clear that her method wasn't magic, and it was always vulnerable to this. While it avoided pitfalls others fell into in other cycles, as soon as there was a really significant split which wasn't substantially correlated with her sole controls of age, sex, and district, her methods were going to yield results divorced from reality. It's a shame it happened so dramatically when she was planning to retire anyway, ending her career on a sour note.
> I think if there's an aberrant thing here, it's people being overimpressed by her track record - overcorrecting based on a small set of dramatic events, starting in 2008
I don't know if they were overimpressed by her track record, so much as they were overimpressed by the fact that she was telling them what they wanted to hear. I swear that 99% of the people telling me how reliable Selzer's polls were on November 2 had never heard of the name Selzer on October 31. She went from obscure specialist to household name and back again faster than anyone since... well, I dunno, I'm sure there's other examples but I've forgotten them all.
There were outliers on the other side too which didn't get nearly as much attention. I remember a few polls showing Trump winning New Hampshire, New Jersey, New York.
Prior to this election, at least Nate Silver was impressed with her, and his judgment counts for something in my book. (A lot of the commentary around the latest poll specifically may have been Democrats hyping news they wanted to here, but readers of Nate Silver would have been familiar with Selzer for a while.)
Yeah I expect Nate Silver to be familiar with Ann Selzer, I was just bugged by all the random civilians who were suddenly telling me that Ann Selzer was the bestest pollster ever despite the fact that they hadn't heard of her a week earlier.
Someone, I think maybe Silver or a commenter on his site, mentioned that the crosstabs were crazy. Way too many women, especially within certain subfields.
This. The only thing I'd add is that she's been famous in large part because she was historically willing to buck the trend, and while she got lucky the last few times (and is actually a good pollster), being willing to buck the trend does mean you'll also be loudly incredibly wrong sometimes. So her being so visibly off here is in some ways correlated with her being impressively right in the past.
I was expecting to see people criticize Dr. Oz as a pick, I was not expecting to see someone criticizing him on the grounds that he was supportive of a trans kid on his show.
Watch the entire clip. The "girl" barely remembers being a boy. Then the mother (it's ALWAYS the mother) "reminds" her that she didn't like getting her hair cut and she compliantly gives the "correct" answer that it was because she was a "girl not a boy." Munchausen syndrome by proxy.
I did watch the clip, which is why I described it the way I did rather than the inflammatory way the tweet does. The mother says they didn't know anything about being transgender until the pediatrician diagnosed it, and I certainly wouldn't accuse someone of making up the story for attention based on a one-minute clip of them.
And regardless of what you think of the mother, Dr. Oz just seems generically polite and validating here, in the way I'd expect most TV personalities to be. Were you expecting him to yell at the girl and tell her she's a boy?
There isn't any plausible scenario for a US civil war in the next 6 years, but I'll try:
Trump goes full slash-and-burn against the Federal government: FBI is disbanded; HUD, Dept of Education eliminated, others radically downsized; Federal funding is pulled from all Universities that have any DEI or grievance-studies departments. This creates massive economic dislocation as significant streams of federal money dry up. At the same time AGI is created at OpenAI and they declare explicit anti-administration political intentions as the state of California aligns with them. The administration begins a serious crackdown on illegal immigration and sends National Guard troops to detain them. Coastal cities declare themselves immigration sanctuaries and refuse to admit federal troops. Trump freaks out about OpenAI and considers it a national security priority to have the newly-minted AGI firmly in control of the federal government. OpenAI refuses and California aligns with them. There is open rioting in the streets as political protests grow unruly so Trump sends Federal troops to California to a) restore order b) detain illegals and c) secure OpenAI. Democrats win the midterms on the back of a 2008-level recession and rampant inflation. They use their control of Congress to impeach Trump on completely baseless politically-motivated charges. The vote is close but he's removed from office amid rumors of unscrupulous practices by the Democrats (which are actually true - they threaten the families of several key swing votes with violence in a very "wink wink" organized crime sort of way) so Trump refuses to leave office and appeals the impeachment to SCOTUS which rules in his favor. Progressives declare SCOTUS a GOP-apologist kangaroo court and refuse to accept the verdict. Coastal progressive states collectively declare Trump an outlaw and refuse to cooperate with the Federal government. Military leaders are divided along regional lines and bases in progressive states formally refuse to acknowledge Trump's legitimacy. Hilarity ensues.
Disgruntled fanatics on the right side of history suddenly discovered the Second Amendment was in fact right and dropped all their calls for gun control so they could tool up and riot in the streets.
Being leftists, they were of course fissiparous, which is why you now have the People's Independent Democratic Republic militia duking it out with the Socialist Democratic People's Front terrorist cells and both of them making and breaking alliances with the Democratic Republic of All Folx Peaceful Protest (And Suitcase Nukes) Units, themselves considered by analysts to be in danger of breaking up between the All Folx, the Y'All Folx, and the "Damn it, it's spelled 'folks' Grammar Communist Cadre" factions.
The rest of the country just grills, drinks beer, and goes to sportsball games 😁
If its had, trump and vanced, possibly rfk, got hit by an american cruise missile; and even your most feminists pro abortation wine aunt far leftist just yields to the right wing and theres just purges, 95% of the miltrey defects or stands down and there's a constitutional crisis; the right wing is in chaos without a figure head and violence being ovisous action but mostly unplanned. I think there are just elements of the leftwing that drink that much koolaid and are just that dumb. If the left wing starts the war in a big obvious way be short and brutal, the neonazis could probably get segregation, dont be an accelerationist.
In, it probably is a slow build up, both sides maintaining plausible deniability; the american empire collapsed, there been a tax protest for a few years and the finance system of the world is just a mess, the world is rapidly decentralizing, mass suicides, and half the effort is shitposting prosuicide messaging to the other side on a fractured no longer global internet; which is very effective especially after the international spying databases get opened up for blackmail; hard line political segregation is happening, with whatever the current rate of people fleeing California is 10x'ed and *every* state. Abortion clinics get bombed in red states, churches in blue, both in purple.
AI leads to superpowered scissors statements and/or superpersuader effects. As a result, you can get any of these:
* Almost every person hates almost every other person.
* Polarization gets taken to the extreme.
* People randomly get sucked into online rabbit holes and emerge as people with drastically different views and values than they had before.
* AI which can convince people of anything just by talking to them, so the first movers get massive cults behind them, and use them however they see fit. Conversation is acknowledged as dangerous, so communications are shut down, exacerbating previous disputes.
Big protest of something or other, police show up to break it up, shots are exchanged, and a celebrity of some kind is killed in the crossfire, prompting both sides to go all in. Who's the American Francis Ferdinand? I'm going with... Weird Al Yankovic. The assassination of Weird Al Yankovic leads to civil war.
If your definition involves neat lines of blue or grey coated men carrying muskets, then I would think the most probable steps involve invention of an ASI which then placed us in a Matrix style simulation, deciding for its own ineffable reasons that the *nineteenth* century was the pinnacle of our civilization.
The Navajo Reservation decides to secede from the United States.
The US Government doesn't want to acknowledge it, but also doesn't want to step in and violently put it down, so there's an awkward state of technical civil war happening.
Would it mean anything legally for the Navajo Nation to secede from the US? Does anyone know?
I'm doubting that "technical civil war" means anything. I think civil war can only exist if a reasonable person thinks it does or some standard like that.
Well if there were a legal means for the Navajo to secede then it wouldn't be a civil war any more. It didn't "mean anything" legally for the South to secede from the US, or the US to secede from the British Empire, but that didn't stop it from happening.
You can imagine various low-level skirmishes going on that turn it into a more real war if you like, to raise it above the threshold where you could say "hmm I guess this does really count as a civil war" but not above the threshold where the US military simply comes in and stomps everyone.
I don't think this is likely, but a tiny civil war that just counts as a civil war seems like a much more likely scenario than a massive war of all against all.
I'm going to say "Trump refuses to leave office and gets a few generals to agree with him on that", or "After Trump leaves office the Democrats charge him with treason and execute him, leading to a popular revolt", or "Congress passes an Australia style firearms ban." But honestly, I can't see any of those as being likely to happen.
Of course civil war isn't likely. But given that it's not likely and given that it happened anyway, I'm curious what sort of scenarios people think are most likely.
Has anyone done a Scott Alexander style “more than you ever wanted to know” about seed oils?
My prior is that the whole thing pattern matches to every nutrition fad: some interesting but not terribly conclusive studies challenge conventional wisdom (in this case, that linoleic-acid is good for heart health) and trigger a current year bogeyman (inflammation!), then pop culture health influencers amp it to 11 and insist they’ve discovered the one magic bullet that explains all our health woes. Bonus points if you can blame it on modernism and double bonus if you can spin a conspiracy theories about BIG FOOD and their effort to POISON us for PROFIT. Until a few years later none of it really pans out, nutritionist update slightly to “maybe eat a bit less of that if you’re sensitive” and the health influencers move on to the next demon food.
But I’d be happy to approach on objective summary with an open mind.
Heather Heying and Bret Weinstein have a sort of intro-level explanation of why seed oils are bad, or rather, why one ought to infer at a layman level that they're probably bad (they presumably have empirical research that supports the layperson reasoning): seeds, unlike other parts of a plant, do not want* to be consumed. They want to grow into new plants. To help with that, they make themselves non-nutritious to animals, relative to, say, fruit, which *does* want to be eaten in order to carry the seeds elsewhere to spread the plant's lineage. In general, we can expect a part of a plant to be nutritious if (a) eating it would help the plant's lineage and (b) the plant wasn't artificially engineered into "forgetting" what helps it.
They have a book called _The Hunter Gatherer's Guide to the 21st Century_, that I think goes into some detail on this.
*"want" here is in the evolutionary sense (Heying and Weinstein are evolutionary biologists) - shorthand for "evolve in the direction of".
"Consumed" here means in the sense of being dismantled and used for nutrition. Simply passing through a GI tract doesn't count toward this. (Indeed, an animal won't want to eat things simply to pass undigested through it, unless those things come with some other benefit. The plant, by contrast, would be perfectly fine with this, and encourage it for its seeds.)
Evolution is too wacky to consistently apply anything like this. There are all sorts of healthy and edible plant parts that aren't the fruiting body; carrots, onions, potatoes. In fact, the fruit of the potato plant is poisonous to humans while the tuber is not, in direct contradiction.
I agree, but will also broaden it in that people aren't the only beings to eat fruits and vegetables. For example, deer eat and enjoy poison ivy, and many birds love eating seeds of various kinds. And anything produced by a plant, if not sprouting into a new plant, will be recycled by other living beings, such as bacteria and fungi.
Well, I agree evolution is pretty chaotic. To take your examples: all three of those vegetables are very aggressively bred from the original wild stock, so to some extent, all bets are off. Or alternately, we say these lineages have "adapted" to being spread by the humans who are functionally symbiotes, and indeed, the human-bred varieties grow in vast numbers.
The wild carrot, aka Queen Anne's lace, birds nest, et al., is a wildflower with a woody root. The seeds are apparently edible and aromatic, but maybe only if cooked, and I can't tell if they're nourishing.
The onion is virtually only existent as an artificial cultivar; the wild varieties (there are several) have bulbs that animals will eat, but the bulbs are not the seeds.
Potatoes grow from their eyes, which are nodes in the tuber. Aha! ...But an animal would have to dig for that tuber. Certainly, they do. I'm honestly not 100% sure what's going on there; maybe the eyes aren't digestible, so gophers et al. end up just spreading them the same way tomatoes get spread.
Hmm. I've eaten onions and other edible stems like leeks growing in the woods, which afaik are wild cultivars. But when eating a fruit, whatever eats it also ingests seeds which are excreted later, propagating the plant. That doesn't work for tubers or stems, since eating them prevents the plant from sprouting. Potato tubers sprout from eyes, which contain the same inedible toxins as the potato fruit. But this can also be avoided by just eating the potato before the eyes sprout.
An animal doesn't want to die, but doesn't much care what happens to its fatty tissues after it does. Either it has surviving kids or it doesn't, and evolution is mostly going to work on "don't get killed until you've had plenty of kids".
A plant, evolutionarily speaking, *does* care what happens to its seeds after it dies. Or even while it still lives but the seeds have fallen/been harvested. So there's reason for evolution to work on "please don't eat my seeds".
Whether this effect is strong enough to have rendered seed oils subtly unhealthy for human consumption. is questionable. Evolutionary biology is quite good at conjuring just-so stories that are probably directionally true, but not at a significant level. And I'd expect that if evolution *had* significantly optimized for "please don't eat my seeds", the result would be something dramatically different than "really tasty oil that probably gives you heart disease in twenty years".
I think the idea there is that animal fat is serving a role for the animal that grows it. Ideally, an animal wants to be as tasteless to predators as possible; however, it also wants to survive times where food is hard to find. Fat is the easiest way to do that, and there are other ways to avoid predators, so animals grow fat, but also claws, quills, sensitive ears, legs that move them fast, etc.
Properly, we ought to apply the same idea to plants. I suspect a seed can only be so unnourishing before it can't even make a new plant, so that's the floor, and the plant has to explore other methods of surviving to the next round as well, and sure enough, while they haven't tried running away, they do grow spines, poisonous leaves, bark, and other tricks.
So then my question would be whether any of these tricks are enough to make a plant just not "care" how good its seeds are to eat, but that ends up functioning on whether it costs energy for a plant to defend its seeds that could be instead channeled into something else that helps it thrive. I imagine a plant wants seeds with enough stored energy to grow that new plant, but not so much energy that animals seek it out, so no matter how I slice it, it makes sense to me that a plant would send some of that energy into animal-repelling tech, but maybe not as much into parts of itself that aren't quite as critical - or even sending that into something that *attracts* animals that will help it spread, such as birds and bats and insects.
I don't know their account much beyond what I described above, but if I tried to lay it out based on what I know of evolution:
Plant seeds will be nutritious or not. If they aren't, then animals will adapt to not eat them, and the seeds will survive, and so will the lineage. If they are nutritious, animals will adapt to eating more of them until they go extinct, or adapt to be non-nutritious or repel the animal with something that leaves that animal worse off, encourages the animal to not digest the seeds, etc.
Ergo, plant seeds "want" to be less nutritious. Ergo, any plant that's evolved on this planet sufficiently long (and hasn't been extensively fiddled with by geneticists or breeders) will possess seeds that aren't nutritious - or are as non-nutritious as the plant can manage while still permitting germination, or will be coupled with toxins.
It isn't one sided though - any animal that wants to eat seeds will be more successful if it can digest seed oils. And humans in particular seem to be pretty good at eating a wide range of plants that have evolved to avoid getting eaten (tolerant of caffeine, capsaicin, etc.)
I think the idea here is to distinguish which things humans have adapted biologically to eating, from which things they've adapted to technologically (either humans to them, or them to humans). The evolutionary biologist's argument is that the former process enables us to be nourished and not poisoned by what we're choosing to eat through good old fashioned natural selection; the latter goes too fast for that to work, and we could easily damage ourselves faster than the biological adaptation process can do its work to acclimate.
One possible answer to this is to speed up acclimation with technology, too. I haven't seen Heying and Weinstein address this possibility specifically, but based on my experience following their Q&A sessions, I'd guess their answer would be that this is possible in theory, but currently, technology doesn't know how to do this. Moreover, they might maintain that the human body, like any lifeform, is so complex that it's likely to outpace any technological solution in the foreseeable future - including the parts of our bodies that get harmed by technologically engineered food. (This is consistent with a rule of thumb I *have* heard them say, which is, when in doubt, the fewer the processing steps performed to make your food edible in the short term, the better it probably is for you in the long term.)
Hmm. Fruit is the only part of plants that is “designed” to be eaten (plus or minus nectar to attract pollinators), and most herbivores do not exclusively eat fruit. Ergo I don’t think the logic fully holds.
At any rate, seeds must contain some degree of nutrition, else they would not perform their function of fueling the newly germinating plant.
Lots of seeds are not easily digestible (to survive being eaten with the fruit) but that doesn’t mean they are necessarily inedible, nonnutritious, or toxic.
The seeds I think of as indigestible are the ones with a hard coating that looks designed to survive going through a GI tract or just being avoided by a rummaging animal. So, yeah.
I agree seeds have to be at least *somewhat* nutritious - while you responded here, I was saying as much to other replies. The idea there is that there are multiple ways for that plant to defend its seeds. One is to reduce the reward to the animal by reducing the nutrition. Another is to raise the cost with seed coatings, and toxins.
In this case I agree with Nassim Taleb, considering that we have a lot of oils we've been using for millenia, why become a guinea pig testing new stuff? it's almost impossible to prove that modern seed oils have no long term adverse effects
The problem is that it's almost impossible to prove that [insert words here] have no long term adverse effects. If this is our standard nothing new can be ever done.
Not at all. If something has been consumed for millenia we can observe long term effects. Also one could argue that genuinely harmful substances would not pass the evolutionary filter but I'm less sure about this argument.
Basically agreed. If something has very visible adverse effects, and it is very obviously linked to consumption of a given food (e.g. group A and B are identical except group B consumes one extra food, and regularly keels over ten years before group A) then maybe...
Generally, though:
- We only started to look at disease patterns systematically with John Snow's work on cholera in 1854. Earlier than that - well, people _were_ dropping dead quickly from cholera, it _was_ from water (from a particular source) that they consumed, and no one isolated the cause before then.
- Even for things we have been consuming for millennia, the dose may change. Sucrose has been consumed for a long time, but the _amounts_ that e.g. the USA population consumes are recent
- It is damned hard to isolate the effects of _one_ food (short of having people drop dead immediately after consuming it). Given different groups which have all sorts of differences, including many differences in diet, attributing illnesses (particularly long term illnesses) to particular causes is excruciatingly difficult
Even more damning, the Romans used lead for plates and plumbing for - at least - centuries, clearly unable to detect its toxicity. Like you said, only obviously toxic foods where the link between consumption and illness is immediate, are easily identified.
I can't add much to what Michael Watts wrote below, that's pretty much what I think
On a related note, while answering this I learned thanks to wikipedia that "naturalistic fallacy" is also a philosophical term that has nothing to do with appeal to nature as an argument
Perhaps I was being imprecise to call it the “naturalistic fallacy”, but whether it’s that, an appeal to nature, an appeal to the precautionary principle, an isolated demand for rigor, or whatever, I think the essence is the same - you’re just stating a prior to prefer more “natural” stuff absent (impossible) proof of the absence of harm from newer less natural stuff. And I was hoping for a more rigorous evaluation of the “seed oils are bad” claim.
Besides, you’re a guinea pig either way because you don’t eat or live like we have “for millennia”. This whole thing started when scientists thought they had (and maybe still have) pretty good evidence that a reliance on saturated animal fats plus the modern lifestyle does have adverse effects - the “for millennia” safety doesn’t apply because we mostly ate a lot less and got killed by other stuff before heart disease from too much tallow would have been a big problem.
Yes, I'm a guinea pig in many ways, but it doesn't mean that I want to experiment with seed oils when a perfectly good time-tested alternative in the form of olive oil exists in my neck of the woods
You're right that it's a prior. But if you want to argue that it's wrong, perhaps you should show that there are numerous traditionally used foods that turned out to be harmful (because we know of many modern ones that were later recognised as harmful). The only one that came to my mind was alcohol, Claude added betel nuts and agave syrup but when I pressed him on it he had to admit that agave syrup is also a recent invention. So it looks like the only undeniably unhealthy traditional foods are those that have psychoactive effects.
>But if you want to argue that it's wrong, perhaps you should show that there are numerous traditionally used foods that turned out to be harmful (because we know of many modern ones that were later recognised as harmful). The only one that came to my mind was alcohol, Claude added betel nuts and agave syrup but when I pressed him on it he had to admit that agave syrup is also a recent invention.
What's the relationship between "considering that we have a lot of oils we've been using for millen[n]ia, why become a guinea pig testing new stuff?" and the naturalistic fallacy?
The assumption that we should be skeptical of new stuff because it’s new (and by implication, leas natural) and that we are “guinea pigs” if we consume it is pretty textbook naturalism.
That's in your mind. The argument is perfectly sound as stated - things we've been doing for thousands of years are unlikely to hurt us. There's no assumption about what is or isn't natural.
Oh, well if you declare it to be correct, how can I object?
Life has a 100% fatality rate. Objectively lots of stuff we do and have been doing a long time is harmful. Objectively lots of stuff we haven’t been doing for thousands of years is helpful to our health and longevity.
At the heart of it, despite the process for getting there, there is nothing really novel about the constituents of seed oil - just the relative composition of the various fatty acid types. Linoleic acid, the supposed bogeyman, is literally essential to human life, we’re just eating *more* of it than we have historically. In a more general sense we *have* been eating heavily processed seeds (cereal grains) for millennia, literally since the birth of agriculture.
So I guess a steel man version of “this is a major change in our historical lipid balance, proceed with caution” I could buy. But that’s a *prior*, not evidence (and the actual evidence seems to show a poor global correlation between seed oil consumption and obesity rates - obesity is much more strongly correlated with GDP).
To the extent Scott has something of a soft prior for Chesterton’s fence barring other evidence, and a general agreement that something about “processed food” is bad, fine (although he seems to share my frustration that nobody can really pin down scientifically what that is). But his article on saturated fat (linked in this thread) is ultimately pretty dismissive of key aspects of the seed oil theory.
> and trigger a current year bogeyman (inflammation!)
This is such a weird bogeyman. Inflammation, like fever, is the reaction you have to something going wrong, your mechanism of putting it right again. The expectation for artificially reducing existing inflammation has to be that it will make your situation worse, not better.
Steelmanning, I presume that the argument is that we want to remove the cause of chronic inflammation, not that we want to suppress the inflammation as a symptom.
In the specific case of seed oils, Don P. is correct. The theory is that a component of the oils (specifically linoleic acid) is a direct cause of inflammation (by one of its metabolites I think?). Basically, “stop consuming this thing that your body feels the need to attack”. Nothing to do with inflammation suppression per se.
You seem to be commenting on a much more general trend.
That's definitely untrue; all the energy is going toward suppressing inflammation itself.
I've asked my mother (a doctor) about this question for fevers, and her take is that we suppress fevers because, in the event that the fever was actually necessary, we can fall back on hospitalization to deal with whatever the fever was going to prevent.
For inflammation, I don't think even that much thought has gone into it; nobody knows what the inflammation is responding to.
Followup question: let's suppose I accept that seed oils might be bad and are probably worth eliminating just in case. What oil should I use in my cooking instead? Olive oil has a low smoke point, and things like coconut or avocado oil have a strong taste.
My experience is that avocado oil has an extremely mild taste, certainly milder than olive oil. Speaking as an avid home chef, avocado oil is my go-to daily oil. For certain use cases, I also keep olive oil (basically just for Italian), bacon grease (richness and searing steaks) and peanut oil (high heat for deep frying) on hand.
Careful! If the pigs ate a lot of seed oils in their feed then there lard will be full of the same kind of fatty acid chains. Ruminants, on the other hand, reprocess everything they eat into new fatty acids. So tallow is safe, lard not necessarily.
Depends on which theory you subscribe to about why seed oils are bad. If you think unsaturated fats are bad, go with lard or ghee. If you think PUFAs are bad, go with safflower oil… a seed oil. If you think a low omega 6 to omega 3 ratio is important, go with canola oil… a seed oil. Uh oh.
As far as I can tell, anything with high omega 3 is going to have a low smoke point.
Lard is very unsaturated compared to butter, ghee, tallow or coconut oil.
Safflower oil is like 70% linoleic acid, a PUFA, and widely regarded to be the most dangerous one. I'm not sure why you think it's low PUFA.
Most seed oil avoiders don't care so much about the omega 3 to 6 ratio, mostly they try to avoid linoleic acid and other unsaturated fats full stop. Butter is what most people cook with.
I’m seeing conflicting versions of fat breakdowns on safflower oil. I was going off a graphic showing it as having a very high percentage of monounsaturated fat.
I think what’s going on is that there are recently (GMO?) high oleic acid varieties of safflower oil being produced.
In any case there does not seem to be a 1-1 correspondence between “seed oil” and “highest linoleic acid” content, at least among plant derived oils, though a couple of the highest linoleic acid varieties are indeed from seeds. Which I thought odd.
Oh, and you’re right about lard relative to beef tallow, but the poster I’m responding to rejected avocado and coconut oil as being strongly flavored, and lard is more neutral tasting IMO.
It mostly focused on high-saturated fat diets, but as a part of that addresses polyunsaturated fats (which mostly come from seed oils in western diets, as you likely well know) as a potential cause of obesity and illness.
Thanks. I have a vague recollection of actually reading that post when it came out, but since the current obsession is all about “seed” oils, I guess it didn’t trigger my memory.
Is the closest, it basically does a high level literature review of studies in seed oils vs not and essentially takes them at their word. It's likely not as good as something from Scott, since I think it's not trying to figure out every possible interpretation from data / figure out if there are other confounders, but I basically trust this like 10-20x more than people making assertions about "the science", which I imagine is what young otherwise be getting.
Has anyone here used Duolicious? I'd expect it to be "just like every other dating app, except possibly even more of a sausage festival," but correct me if I'm wrong, please.
I tried it. Girls there were a lot more internet culture, which is good if that's your thing. It was also so much of a sausage fest that I hardly ever got replies. Much lower response rate than other dating sites. Chicks mentioned having thousands of messages to get through. This was closer to launch, so maybe the situation has improved since.
I took the SAT four times in my life and remember each score I got. I also remember my GRE score.
If you ask someone what they got on the SAT, ACT, or GRE, and they took it fewer than 10 years ago, but they respond they can't remember their score, how likely is it that they got a bad score and don't want to admit it? 50% likely? 90% likely?
Data point: I got a very high score (1598) and my default in conversation is to claim that I can't remember. Sharing a score that high is interpreted as bragging and can get very awkward.
I don't think I've ever been in a conversation where people talked or asked about SAT scores. I remember mine, but I never tell anyone because I assume no one cares.
I wouldn't take that as evidence either way. I personally barely looked at my scores bc I knew I'd get a good score. Like GlacierCow, I do remember my percentiles though. An obviously better metric, in my opinion and way easier to remember.
Something I've been asking about for a while, from that link:
> The Biggest SAT → IQ Conversion Site Is Wrong
>> This man has directly converted the SAT percentiles to IQ scores, which is not what should be done. Tests like the ACT and SAT correlate with IQ at about 0.8-0.85
>> Instead, the ideal way to do this is to take the percentiles from the current versions of the SAT and then convert those into z-scores and then regress those z-scores by the mean by the estimated regression coefficient.
I don't see how this can be correct. When we see that the SAT correlates with IQ tests at about 0.8, we also note that that is the same correlation we see between different IQ tests. But we don't regress the results of other IQ tests when reporting the tested IQ given by those instruments - if you take a Wechsler test and your reported IQ is 120, then your tested IQ is 120 regardless of whether I'm secretly thinking about Wechsler or Raven's. Why would we regress results only for the SAT?
How is that a direct answer? It tells you literally nothing about the question you asked. It does tell you that people who don't remember their SAT score are stupider than people who do, but that doesn't speak to the odds of forgetting versus not wishing to report.
I don't remember my score exactly (been more than 10 years since then) because I don't even know what the scoring standards are anymore (various SAT scoring system changes over the years). But I do remember that I scored in the 97th percentile and I remember this because it's the closest analogue I have for IQ I have. What does this make me?
2) Do you have absolutely no memory of what your score was, or could you confidently narrow it down to something like "around 1400" or "in the low 1300s"?
If I want to guess it was something like 750 on math, a little less good on the reading/writing. Looking online I was in the 2400-score-years so it gets confusing. I think I could confidently narrow it to something around 2100 (+-100ish?) but I had to look online just now to remember what that meant so I think I've contaminated the example.
Yes but do you remember the fastest time you ever got on a 100m sprint?
People who base their self worth on test scores are likely to remember, people who don't are more likely to forget.
Why would you even ask someone that, if you're not in charge of college admissions or something? Just so that you can feel superior if their score is lower than yours? I don't blame anyone for weaseling out of that sort of game.
I guess one of the first things Ted Cruise asks someone he meets for the first time is, “What is your IQ?, if they don’t know he asks “What were your SAT scores?”
It depends on how many bits of information are involved? I also remember, even though it’s been a lot more than 10 years, but it’s not hard since I only have to remember one bit of information.
But also, why do you care? I’ve never had anyone ask me my sat score after high school, and I would definitely judge negatively someone who asked me that today.
I took the PSAT and the SAT, once each, and remember both my scores but not which one goes with which test. They're pretty close, but less than perfect, so it doesn't much matter.
That's not an approach I would recommend when the scoring scales are so different in absolute terms. It's similar to suggesting that one mile plus one meter is two. You'd need to convert both scores to the same scale first.
I also wouldn't recommend it as stated even for multiple instances of the same test; your compressed score will get worse and worse the more you take the test.
One thing I don't understand is why it's constitutional for Trump to run for office in 2024. Amendment 22 says you can only have two terms, and Trump won in both 2016 and 2020.
I guess on a serious note, Trump running for office in 2024 is an implicit admission of the loss in 2020. That is, if we assume Trump cares about the Constitution. Well, never mind!
Are you Attorney Paul Gordon, submitter of the hilarious legal troll challenge making that argument to the Maine Secretary of State? That objection was considered and rejected: "Application of the term limit turns on whether an individual has _actually been elected President twice_, not on beliefs or assertions about that fact."
Interesting, the plain language of "No person shall be elected to the office of the President more than twice" suggests it's the election that matters, not the ratification.
The election by the president happens by the electors, with the election of *those* electors being a separate matter, insofar as I've understood. The electors conducted the election of the president on Jan 6.
When precisely do you believe the election is COMPLETE, and the person is considered to be elected? When the last polls close on election night? When the ballots are all completely counted? When ENOUGH ballots have been counted? When all recounts have been completed? When appeals/legal challenges have been exhausted?
Or perhaps all of that is irrelevant, and it's when the electoral college meets that matters, which was December 14 in 2020? (That one's actually reasonable, and Trump lost that one.)
I don't believe he has issued any official statement about it. But also, it's the certification on January 6 that determines whether or not he (retroactively) lost on December 14. At any rate, I don't think the precise time one officially wins is something anyone has cared to adjudicate, since the inauguration (specifically the moment he completes swearing the oath of office) is generally what matters.
I don't know where I originally found the link, but this guy Sam Jennings has the best take I've seen on Trump's victory. A few extracts:
> Which is not to say I'm excited for the coming chaos. Only a very certain, or very rash, person would prefer unbridled chaos to an anxious blandness. And yet we have no real choice in this. The people have chosen chaos. They’ve hit the big red button that says, “Go on, fuck it all up, what do we have to lose?” and now we’ll get to see exactly what we have to lose.
[...]
> It’s an awful picture, yet I can think of no other way to put it: the entire attitude post-election seems to me to be essentially post-ejaculatory. We’ve been anxiously invested and secretly turned on by a carnivalesque libidinal ritual, being hacked out by two very unsexy parties, and now we’re all emerging from a vaguely masturbatory haze to discover how sordid and sad the whole thing was.
> and now we’re all emerging from a vaguely masturbatory haze to discover how sordid and sad the whole thing was.
For some? Sure. For Trump's core base? Smells like projection. People like to pretend that Trumpism is just some inexplicable wave of hysteria. When in fact, the stage was set 30 years ago [0]. Keynes once said "When the facts change, I change my mind. What do you do, sir?" Well apparently, Jennings' go-to strat is to revert to epistemic helplessness. It's not as bad as full-on denial, I'll give him that.
Say it with me, everyone: "We're not at the end of history"
My favourite take was that Kamala Harris ran a campaign about ideas, and that non-college educated people voted as a bloc for the person whose campaign was about them.
I don’t think everyone who voted for Trump just wants to watch the world burn, as noble as that is.
Lots and lots of people who voted for Trump but want opposing things think Trump is going to be on their side. They can’t all be right! It will be interesting to see how it shakes out.
How is it unfounded ? It is very likely that Harris would have kept on with the current policies. It is also very likely that she wouldn't have tried to break any norms.
I keep seeing people trying to "both sides" this so hard but it always seems so dishonest. For example equating Hillary's "[Trump] knows he’s an illegitimate president" which is calling out shady voter suppression tactics and the fact that Trump did not win the popular vote with the staunch denial of the 2020 election result by Trump to this day and the organization of a (failed) plot to remain in power.
Are you people serious ? Do I really have to convince people in the comment section of a rationalist author that Kamala Harris was willing "to destroy the country" ?
"It is very likely that Harris would have kept on with the current policies. It is also very likely that she wouldn't have tried to break any norms."
I agree with this 100%. That said, I have a different interpretation: if you keep on with the current policies, and you don't break any norms, the country is going to be destroyed. And she's "willing" to let it happen. She doesn't think it *will* be destroyed, of course, but that's where we disagree.
Well at least we share the same reality but if you think keeping on the tracks of the previous administration is going to destroy the country you may as well define what destroy means to you because it seems far off the center of even the metaphorical meaning of this word.
Yes, they are; but no, you don't. I put it to you that there are better uses for your time than trying to interrupt people who are deep into their enjoyment of a gratifying fantasy that, while absurd, can never truly be debunked.
The Harris administration will always be a counterfactual. You can't even prove to *me*--a person who thinks it's broken-brained delusion to believe otherwise--that President Kamala Harris would have continued to be a bog standard bland Democratic politician, presiding over a bland (but diverse!) Democratic cabinet and pursuing vaguely center-left policy objectives, to the extent possible with a divided Congress, in the usual boring and bureaucratic manner for four to eight years. How do you expect to prove it to people who genuinely get a kick out of imagining her becoming Woke Hugo Chavez overnight?
"You can't even prove to *me*--a person who thinks it's broken-brained delusion to believe otherwise--that President Kamala Harris would have continued to be a bog standard bland Democratic politician"
Look at the record. She got a reputation as "Copmala" in California. Later on, when first running for the primary, she tacked hard left because the winds of change had shifted to a different position being considered more appealing to the public. That got her nowhere, so this time round she moved back to the centre and avoided explaining why she was repudiating her former positions.
That's because she doesn't *have* any positions. As Vice-President, she's best known for what? A media controversy over whether or not she was known as/appointed as "Border Czar"? Her campaign this time round ran on vibes.
If elected, she would have continued to be as bland as rice pudding and would just have gone along with whatever seemed to be the best position to give her the greatest chance of high approval ratings. She wouldn't have gone full socialist and she wouldn't have reverted back to tough on crime (legal weed and free loans for black men to transition from being drug dealers to being legal drug dealers/businessmen still has me face-palming here).
Yes but that's a good thing in a democracy. You want someone who will do what's required within the limit of the system. Always going the in the direction of the person in power's latest whim is not a good thing. It strains the system and may end up breaking it. I'll tank blank rice pudding instead of castor oil any day.
>In 2019, Vice President Kamala Harris told CNN’s Jake Tapper that social media companies “are directly speaking to millions and millions of people without any level of oversight or regulation[censorship] and _it has to stop_.”
[emphasis added]
And, in addition to her words, she was a participant in the Biden/Harris administration, which used Facebook and pre-Musk Twitter as private proxies for unconstitutional censorship by their administration. So this is not just a hypothetical of what Trump _might_ do.
There's an argument that she was pro-censorship merely because that's what the polls suggested her voters wanted. If the polls said her voters were 60-40 in favor of tariffs on quinoa, Tapper would have been hearing her rail about how Peruvian agri-conglomerates "have put thousands of small American quinoa farms out of business without any remedy and _it has to stop_."
I took this as Deiseach's point: Harris lost appeal because she resembled a weathervane. Even her speech stance might have been poll-driven. (I think it's likely, and all it would have taken to flip her would be a minor BLM tweet storm.)
It occurs to me that you never actually /hear/ this sort of thing any more. It's too gauche to mudsling out loud, even at the end of a night's drinking. Mostly, now, you /read/ it in the more shouty bits of the internet.
...What's so horrible about that ad? That seems like a completely run-of-the-mill political ad. I thought you were going to link something calling Trump Hitler or something, but no, it's just making the obvious point that Trump is not aligned with the working class in any capacity.
Ah, it's that the guy is obviously An Actor playing the part of "regular guy who goes to football matches with his dad". Sure, a real Regular Guy would have been stiff as a plank in the part (if you've ever heard radio ads with Real People, you'll know what I mean). So it's plastic not real wood, while the tone of the ad is "we're real, he's not".
Harris' disdain for the common man bleeds from it. There's the prevailing sense that she thinks people are cows and they'll follow whoever looks most familiar. "He lives in a country club, that's not your herd! I'm your herd, with the haircut and the Sports!"
This is a twisted and uncharitable take. Do you really think Harris personally oversaw this ad ?
A lot of ads I see from the US country are like that : Interview someone that is supposedly just one of the viewer so that they identify and buy something. Nothing special about it.
No, I doubt Harris had more to do with it than "my campaign signs off on these". It's a terrible ad, but no worse than many political ads, so she can't be blamed for that.
But "Trump is a fake!" messaging loses out when the guy spouting it is clearly professional actor doing a role and he'll be dressed up as white collar PMC guy, possibly IT guy, for another ad instead of "I'm a guy what drives a SUV and work in a saw mill".
Contrast that with the garbage truck, and that's really Trump in a real garbage truck. It may be dumb, but it's funny and most importantly it's not faked up the same way.
Harris is in charge of the organisation that made this ad, her personal level of involvement doesn't matter. Being senior management means taking responsibility for the actions of the organisation that you built, and Harris was asking for the most senior-management job of all.
I'd never seen this ad before. It's not quite as terrible as the "I'm a man what eats carburettors for breakfast" one, but it oozes the same contempt for voters. "We have to figure out the things that each of the demographic groups on rows A through M of this spreadsheet likes, and make an ad filled with each of them, and say Donald Trump Bad. Make sure it doesn't mention Kamala Harris or any actual policies."
>Do you really think Harris personally oversaw this ad ?
It ends with "I'm Kamala Harris, and I approved this message", so yes, she personally oversaw it as much as she was going to personally oversee the presidency.
...I see your point. Still, I don't think it comes from a place of contempt. There is a genuine lack of understanding on why the working class loves Trump so much, fueled by blind celebrity worship...
I guess the reason doesn't matter though. The consequences for failure are all the same, and I have no sympathy for the left's coming death.
The campaign funding reminds me of 2016 when Hillary's campaign was boasting of all the money it was raising and there were constant media pieces about "Trump has no donors, no money, his campaign is going to run out of steam soon".
Well, looks like the deeper pockets didn't win a second time! And Trump's campaign didn't need to spend money on publicity because everyone was throwing free publicity at it with both hands! From "The Republicans are weird" to every breathless "it's the Fourth Reich for sure this time!" social media/traditional media take, they honestly spent more time talking about Trump than their own candidates.
What was Walz gonna do in the White House? I have no idea, but I certainly know the Harris campaign position on JD Vance.
The difference this time round seems to be the begging letters/emails being sent out still, and in increasing numbers. The over-spend seems to be about $20 million, which is not terribly bad, but the reaction makes me wonder if there is more there we are not hearing about. Is the DNC growing a spine and demanding Harris personally take responsibility for paying it off or making sure it's paid off? Are the deep-pocket donors cajoled into backing the booting out of Biden now having buyer's regret and intimating they may not be putting their hands in their pockets the next time a Democratic candidate comes looking for a little help to protect democracy? Reading between the lines, there seems to be some indication of the usual bribery re: walking around money to get out the black vote, that may not sit well with the emphasis on transparency and openness and 'Kamala was the grassroots choice'. Also that consultants, hangers-on, and 'jobs for the boys' chomped a good bite out of the funds, but that's par for the course in politics.
"Harris camp also spent a whopping six figures building a set for her appearance on the “Call Her Daddy” podcast, which attracted a fraction of the viewership of Trump’s appearance on the “Joe Rogan Experience.”
Yeah, that sure was worth the money they tossed down the sink, wasn't it? This kind of thing niggles at me - the campaign seemed to want a huge level of control over such interviews (not going to the studio where this podcast is usually held, and the same with what Joe Rogan has said about them - they wouldn't let her go to him, they insisted he had to go to them) which sounds like they didn't trust their own candidate enough to let her do an unscripted interview on strange territory. That doesn't bode well for "she is competent and capable".
There seems to be an air of desperation that I don't remember from Hillary's loss this time round.
I think I might disagree with him in that in many ways the grift is the point for a lot of Dems; I think my best steel-manned version of this "functioning as a political party that wins elections is secondary to its function as social club for the Professional Managerial Class and above, and throwing out large contracts for bad services is merely a cleverly-disguised welfare payment
To that last point - from my recollection, the atmosphere in November 2016 was complete shock - "how *could* this happen" overshadowing "how *did* this happen", with blame focused more on Americans failing to elect Hillary than her failing to appeal to them. The latter came later.
A couple of thoughts. One, I don't think the people who voted in Trump are having any kind of second thoughts or "post-ejaculatory" worries. They seem pretty content with having won and are looking forward to seeing how far Trump can get with fixing their problems. I don't think many, maybe any, have delusions that everything will go their way, but they have a reason to hope that some will go their way. Second, and this is important for people who hate Trump to understand, Trump only exists because the people who support him have felt like they didn't get *any* of their preferred preferences since sometime in the 80s or 90s. NAFTA and greatly increased immigration have done huge damage to them economically and socially. If the choice is "more of the same" or "unbridled chaos" they pretty much have to pick chaos. It's the choice between some level of hope and death.
It could be true if conservative living is just that much better than liberal/left. That's not a charitable option, but there are so many possible confounders that I don't know that we could ever really determine the reason(s).
My personal take is that rural living really is that much better for human happiness, and rural people are significantly more conservative. Less crowded, more likely you know your community members. Less chance of running into random jerks and criminals, even if the rate of criminality/jerkiness is the same or worse in rural areas. There's something to be said about the ability of one jerk to ruin your day, and the chances of running into one jerk a day is unbelievably higher in a city (unless the jerk is your family or a close neighbor or something).
The first is that the conservative (small c) life script is genuinely better for human flourishing. Finish school, get a job, get married, have kids, go to church and engage with your community and extended family (all while living in the countryside or suburbs).
Compare with 'Finish school, do a postgraduate degree, move to the city and try to break into a competitive field, sleep around for several years, cohabit and then maybe have a child in a small apartment without getting married'.
All the elements of the first option are better for psychological health.
The second element is that the conservative worldview itself is better psychologically. Believing that you are responsible for your own fate, believing that the world is basically just, believing that you have duties to other people and to ideas bigger than yourself are all psychologically helpful.
Jonathan Haidt has done some interesting work on this. He refers to the modern leftist worldview as 'anti-CBT' for this reason.
I don't really agree with most of their policy preferences, but society has been running on "Certain people have gotten their way too long and don't get to have their way anymore" for so long that half of the members of "certain people" haven't actually seen anything go their way for their entire lives - up until Trump. Society has been on this train for about thirty years now.
("Certain people" is not "white people", or "men", or "Christian"; it's both more specific and more general than any of those things.)
Yes, I think that the process of being disappointed by the politician you supported is usually a gradual process, not a rapid one analogous to post-nut clarity.
I think that Harris supporters might be having a quicker-than-usual realisation that she was shit, but that's only because they all thought she was shit four months ago and then suddenly turned on a dime the moment she became the only possible nominee.
>I think that the process of being disappointed by the politician you supported is usually a gradual process
Well, I viewed the contest as between Horrible and Worse (with Harris _narrowly_ getting the role for Worse), so I'll only _really_ be disappointed if Trump actually manages to stumble into a nuclear exchange (unlikely, but possible - but it was possible with Harris too).
Given that the only other choice was an idiot fascist, you gotta line up behind the lukewarm neolib. I mean, the last lukewarm neolib squeaked his way in.
Who knows though, maybe running a centrist campaign and getting blown the fuck out will knock something loose in the D's hydrocephalitic party apparatus. One can hope.
I got the impression Harris was trying to project centrist and people didn't buy it. After four months of Harris being in the spotlight, I still have little idea of who she really is. The best glimpse I got was during the debate, for which she had a month to prepare, and all I saw was preparation.
People who personally hate trump and generally don't like the paleocon style politics of his hangers on didn't get fired up by:
"No socialised medicine, Expand Drilling, Support Israel to the hilt, I Love Rich People, Here I am standing next to a fucking Cheyne" style messaging, and your actual conservatives and centrists (embarrassed conservates) were never going to vote for a black woman running against their favorite idiot old man on the back of "Look, I know everything is more expensive, rates of savings are the lowest they been for decades, you are one broken leg away from missing rent and being homeless, but line go up!"
"I still have little idea of who she really is... all I saw was preparation"
I never understood this argument, bc it always seem like any criticism you can levy at Kamala would be one thousand times worse with Trump. No, I would never prefer an honest Hitler to a passive-aggressive faux-kind co-worker.
Yes, she prepared for a debate, sure, she isn't being super genuine in her policies, but is that really worse than Trump? Someone who didn't plan at all for the debate, didn't have any policies proposals ready, and lied through his teeth? Trump is a genuinely stupid man, who does not understand much about how the government functions. How is he any better?
People wonder about how a civil war in the US might break out. How plausible does the following scenario sound?
1. Trump purges the army of commanders not considered loyal to him (under the guise of some other reason)
2. Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
3. 2028 rolls around and a Democrat is elected POTUS. He tells the army to withdraw from the cities but meanwhile Trump tells them not to and at least one of the armies obeys Trump not the new POTUS. (Throughout the period after martial law is declared, The Left stays angry about the situation, with violence breaking out occasionally, army personnel sometimes getting shot by The Resistance.)
4. Some generals stay loyal to Trump while other generals are loyal to the POTUS.
I think waiting until 2028 would be a bad play for him. It would give the opposition time to find a figurehead to coalesce around. Shutting down the opposition should happen before then, optimally during or a bit after the midterms. Of course, he would need to artificially inflame tensions domestically; possible options for that are doing something blatantly unconstitutional that is still popular with his supporters, the deportation program getting revealed to have caused deaths or other human rights abuses, or simply enabling Israel to finish the job in Gaza.
Very unlikely. Every stop of your process seems individually unlikely, with a combined effect of all but impossible.
1) Gutting enough of the military to make it specifically loyal to Trump would completely eviscerate the military. Rank and file people may love Trump, but not separate from their love of the US. The commanders don't particularly love Trump.
2) Limited martial law is possible, but not multi-year. The federal government calling for martial law would likely result in local jurisdictions doing everything in their power to fix a problem before federal troops could even get there, and then quickly getting them out. Martial law is bad for everyone involved and would be incredibly unpopular, especially after an immediate need was handled.
3) This seems to presuppose that Trump doesn't leave office? I'm not sure what you have in mind, but once we have a new president who can give lawful orders to stand down, then we don't have Trump in office to give counter orders. If Trump would fight against leaving office then the details of his argument and who supports him matter. I don't see that as being very likely, either.
4) If Trump could pull off the complete gutting of the military, I guess some of the new generals could potentially support Trump. But they would almost certainly be very bad generals, likely incompetent at their jobs and unliked by their troops. If they were qualified people who rose in the ranks after their bosses were removed, then why follow an illegitimate Trump?
>1. Trump purges the army of commanders not considered loyal to him (under the guise of some other reason)
Seems definitely possible. I'd say at least 50% chance this happens anyways.
>2. Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
Very unlikely. Maybe 1-3% chance IMO. There are a lot of other steps before that: declaring national emergency, militarizing the police, sending in the national guard, federalizing the state national guard, repealing or ignoring all the laws that prevent him from doing this...etc. etc. Public sentiment would be low. Congress would likely act against it. Supreme Court is, contrary to popular opinion, not a bunch of Trump loyalists (they're FedSoc which is very much not a MAGA organization!) and would probably act as well. States would likely refuse to cooperate.
>3. 2028 rolls around and a Democrat is elected POTUS. He tells the army to withdraw from the cities but meanwhile Trump tells them not to and at least one of the armies obeys Trump not the new POTUS. (Throughout the period after martial law is declared, The Left stays angry about the situation, with violence breaking out occasionally, army personnel sometimes getting shot by The Resistance.)
less than 1% chance. Highly unlikely
>4. Some generals stay loyal to Trump while other generals are loyal to the POTUS.
less than 1% chance. Highly unlikely.
>5. Civil War
less than 1% chance. Highly unlikely.
I can think of like 10 different possible paths off the top of my head to civil war that are more likely than this. Most involve some kind of state secession crisis or medium-intensity insurgency, rather than a president trying to become king.
My guess would be to sequester them from the general populace, official purpose being "special training". Some of them are selected for special teams, given special gear and instructions for using it.
Generally, any government that wants to oppress its own people gets that going by having a police force that doesn't socialize with those people. Very, very unlikely in the US as a result. The closest we tend to get is hiring police from one part of town to police another part of town.
On "1" there's a big gap between "personally loyal to Trump" and "actively disloyal to Trump". I might expect him to fire a few of the actively disloyal ones, but the vast majority of officers are in the middle, willing to obey all legal orders.
> Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
Cities are wildly leftwing, and quite well armed in practice
Riots happen because they are allowed to happen, if trump is threatening to end them using the national guard and it would be mildly embarrassing, the riots will stop.
> How plausible does something like that sound?
The administrative state donates like 98% of the money they spend on elections to democrats; the idea trump finds people to staff the current state with lackeys is just wrong. Even if he purges(fingers crossed), how will he find the manpower of people trained in the machine without dipping into those who are merely loyal to the paying the bills.
If trump changes the character of the american empire, the american empire is getting much smaller at which point it will be more of a states rights issue, unless you believe trump will control Californians governor.
----
I dont see how a rightwing-Establishment backed civil war happens, its the leftwing using the Establishment vs the right wing rural factions(with you know, guns, food and land; farmers protests are very powerful for a reason) thats ever been on the table, cities vs rural is the fundamental divide in politics and always will be, food needs land, high speed trade need dense population, cities rarely see the farms but are wildly dependent on them.
And the country has spoken, the outsider is invited in to rebalence the scales, he will be surrounded by a bunch of leftwing insiders looking to block him at every turn; this is good for peace(assuming he survives assassination attempts)
Incredibly unlikely. Trump purging enough people from DoD to where they would support long-term suspension of Posse Comitatus to implement Martial Law is one where the Army wouldn't be competent enough to do it anyway.
I don't know if it's been said on here before (and I'd love to hear @Deiseach's perspective on this), but elsewhere it's been noted how fascinating it is that people's mental model of an American Civil War is something that looks a lot like the Spanish Civil War and not "Na Trioblóidí, but bigger". Because I think we're a lot closer to the latter than I think we would like to believe, and I think the former is much harder to make happen than I think people expect, to say nothing of "Nicaragua but with more paramilitary death squads".
The divisions in the US are nothing like Northern Ireland. You asked Deisheach who is, Irish. I’m British with family in Northern Ireland - so with an understanding of history of the place.
These ethno religious civil wars are not what is happening or possible in the US. An ideological civil war is unlikely either, despite the caterwauling about Trump, he is not a fascist and the opposition are not communists either. In the Spanish civil war they were, literally that. That, and anarchists. Democracy was a young political idea then and had little credibility.
In the US it is part of the fabric, however flawed that fabric, for centuries.
I did worry slightly about the US creating, out of nowhere, inter ethnic conflict at the height of the anti white supremacy kerfuffle a few years back, those worries have entirely dissipated in this election. It looks like Hispanic males in particular either like a bit of white supremacism or don’t believe it.
Being old enough to be around during the 70s and just old enough then to be somewhat aware of what was going on in America, I remain astounded you didn't have your own version of The Troubles. You really got lucky and zigged where other countries would have zagged. You had your own amateur terrorist militias robbing banks and performing bombings, you had political theorists pushing revolution, and there is of course the race question.
Seems like your Founding Fathers got something right when setting up the state!
Probably helped a lot that our most dedicated and extreme revolutionaries could always defect and go back to cushy jobs thanks to being rich kids. No need to be a diehard martyr, you can just become an education professor!
Try iodine. If that doesn't help, organic food. Sulfonylureas are used as preharvest desiccants in conventional agriculture. They're also used as diabetes drugs, because they make your pancreas squirt out insulin. Long-term, this can lead to an insulin resistance phenotype, and hypertriglyceridemia can follow on from that.
Im an anti-reductionist; I believe logical positivism was proven wrong a century ago with gobel incompleteness but 99% of people just let it continue to effect their thinking with things like the simulation theory(what if the 3 body problem is unsolvable computationally, there wont be an infite chain of universes being simulated as each simulation wastes more resources)
This argument doesn't seem to work for most people, its to foundational, and maybe its to big a epistemic switching cost? So new thought experiment, the color blind prisoner:
You are colorblind and at the mercy of a warren who is a full-sighted supremacist, he is running "games" to find who to free and who to kill, you being infinitely smart deduce his methods of randomly assigning colors, the flaws of the coins so you have 50.0001% at predicting his coin flips etc. Do you really think you survive? Or do you fundamentally lack the right sensor? Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
It is currently believed in this community that ai will deduce physics of the real world from video: https://www.youtube.com/watch?v=fVN_5xsMDdg why? How do you discover dna without a microscope? How do you discover quantum physics without the double silt? How do you discover trustworthy faces without human feedback?
Critical data exists, and is an "unknown unknown" until stumbled upon; you must poke and prod reality you cant merely watch and learn everything.
I'm not 100% confident I understand. But if you're saying "at some point, you need to stop armchairing and use your eyeballs", that's a theme Yudkowsy has definitely covered. In fact, that's the entire point of his obsession with Bayes' Theorem. More specifically, there's a post that comes to mind where he talks about how syllogisms only tell you about possible worlds, and you need evidence to narrow things down further. I'll edit if i can find it.
> I dont know why 3 of yous have brought up yudkowsy
It's time to recite the lore of yore.
Before Scott blogged on Astral Codex Ten (ACX), he had another blog called SlateStarCodex (SSC). Scott deleted SSC for a few months(?) in retaliation toward the New York Times (NYT). He gave an interview to the NYT about COVID-19, but the NYT used his real name (n.b. "Scott Alexander" is a pen-name) and Scott preferred not to self-dox for various reasons. Before SSC, he blogged over at Lesswrong (LW) under the handle "Yvain". LW is a community blog, and was the epicenter of the diaspora of Bay Area Rationalists. LW was founded by Elizer Yudkowsky (EY), who was obsessed with AI x-risk and Bayes's Theorem, among other topics. Before LW, Scott had a livejournal. I forget his handle. Scott and EY were the two most prolific writers on LW, I think.
The "ai will deduce physics of the real world from video" idea comes directly from EY's posts on LW about Solomonoff Induction (SI), AIXI, etc. More specifically, the Rational Animations video is a roughly paraphrased narration of EY's "That Alien Message" [0].
> Riemann invented his geometries before Einstein had a use for them; the physics of our universe is not that complicated in an absolute sense. A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a hypothesis — perhaps not the dominant hypothesis, compared to Newtonian mechanics, but still a hypothesis under direct consideration — by the time it had seen the third frame of a falling apple. It might guess it from the first frame, if it saw the statics of a bent blade of grass.
In all likelihood, the three of us just presumed you were familiar with the lore without looking at the video, and were knowingly engaging with EY's arguments. Now that I've watched the video, I don't think it did a great job of explaining what the actual point of the story was. EY's claim isn't that AIXI can deduce all of physics from a few video frames with absolute certainty, but that it can at least *brainstorm* lots of plausible theories, based on the assumption that the physics of reality aren't pathologically obtuse, and that AIXI subjectively experiences reality millions of times faster than humans.
(N.b. I have my own opinion of EY's argument. But first, it's important to establish what the argument is, and where it comes from.)
(Edit: for additional context, EY claimed that Bayes' Theorem is a generalization of the scientific method. Bayes' Theorem is useful when you're poking, prodding, and A/B testing when you're allowed to do that. But unlike science, it's *also* useful for times when all you can do is passively observe, like in the case of astronomy. Maybe this helps clarify WindUponWaves' comment.)
> (what if the 3 body problem is unsolvable computationally, there wont be an infite chain of universes being simulated as each simulation wastes more resources)
I don't think that's what people mean when they say the 3-body problem is unsolvable.
AFAIK, the 3-body problem is *chaotic*. "chaotic" means if you make a prediction about a system's output, it can wildly diverge from reality because the measurement-error snowballs out of control. You can certainly simulate it, though. E.g. a pinball machine is "chaotic" in the sense that a tiny change in the angle of its initial trajectory can lead to the ball flying in opposite directions off its next rebound. If you were to try to "solve" the path "analytically" by matching it to a one-size-fits-all shape (like say... an ellipse), you're going to be disappointed. But if it were unsimulable, pinball videogames wouldn't exist.
If you cant simplify the 3 body problem then simulating an electron with say 3 64 bit floats, means you used on the cheap end you used 192 electrons to simulate 1; this stops infinite regress.
Oh, that sounds like an observation about the map necessarily being smaller than the territory (else it's not a map, it's a copy). In which case, mentioning the 3-body problem obscures more than clarifies.
I mean, I kinda see what you're getting at, from your examples. But I suspect you're confused about certain concepts.
A) Logical Positivism embraces empiricism.
B) Bay Area Rationalism embraces empiricism.
C) 17th Century Rationalism *shuns* empiricism. (Many, such as TheAncientGeek, have complained that EY's name for his community was *very* poorly chosen.)
> I believe we are seeing glimpses of a non-reductionist biology and we will *finally* discard to notion that any system evolution creates only does 1 thing at a time with cleanly defined boarders.
D) Uh... I don't think that's what "reductionism" means? Nor is that representative of what biologists believe? Where are you getting this from.
> You are colorblind and at the mercy of a warren who is a full-sighted supremacist, he is running "games" to find who to free and who to kill, you being infinitely smart deduce his methods of randomly assigning colors, the flaws of the coins so you have 50.0001% at predicting his coin flips etc. Do you really think you survive? Or do you fundamentally lack the right sensor? Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
btw, this sounds like the Multi-Armed Bandit Problem, which highlights the "exploration vs exploitation" trade-off. I.e. "is it better to min-max a particular slot-machine, when a neighboring slot-machine might practically print money for you?" I remember there was some discussion in LW's Geometric Rationality Sequence [0] (not EY) regarding the Bandit Problem and Thompson Sampling. In case you're interested in reading or participating.
Yudkowsky is just describing history. He's not even speculating, he's just describing what's already happened, because his claim is a lot more modest than people think.
As in, what he actually said was,
"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a *hypothesis*..."
Not that it would prove it, or deduce that it must be true. Just come up with it, in a "throw shit at the wall and see what sticks" way, like what particle physicists and string theorists do while waiting for the next supercollider or gravitational wave detector to be built. If you read or listen to the original context, it almost sounds like he's describing a fandom coming up with wild ass guesses about what the rest of the series holds, because they're impatient, this is good fun, and just by chance *one* of them is probably going to be right if they cast a wide enough net.
(In fact, my college roommates and I did exactly that, about the physics of a hypothetical universe -- The Prestige, regarding Tesla's cloning device, if you're curious. One of the best times of my life...)
Anyways, Yudkowsky isn't even speculating. He's just describing what's already happened in history. The classic example would be the mathematics of ellipses -- in short, they were invented by the Ancient Greeks for such sterling uses as "describing the funny shape the shadow of a Sundial tip makes as the Sun goes round" and "trying to do something that's literally impossible, a fool's errand" (doubling the cube, a.k.a. squaring the circle, a.k.a. construct a cube that's precisely twice the volume of another cube, using only rational numbers, despite the fact that the cube root of 2 is irrational. But the Greeks didn't know that. What they did know was that ellipses were useless for practical matters like astronomy, because the orbits of the planets are *obviously* perfect circles, as befits their august position in the heavens).
Then, 1800 years later... those ellipses turn out to perfectly describe the orbits of the planets, once you gather more data by inventing telescopes. In fact, they're necessary for things like "Heliocentrism" to work, because otherwise Galileo-style Heliocentrism is *worse* at predicting the orbits of the planets than the Ptolemic Geocentric model, and no one will take Heliocentrism seriously if it's simply worse at the job.
But instead of the scientific community needing a lifetime of work to try out all possible orbit shapes and the math necessary to make each one work... Kepler reads an old book and realizes the job has already been done for them. The lifetime of mathematical work has already been done for them, ahead of time, by someone they've never met. Menaechmus the Ancient Greek mathematician had never looked at a planet through a telescope, and might not have ever actually *seen* a planet in the night sky at all if his eyesight was bad enough... but his work describes the orbits of the planets nonetheless. Not bad for someone who got even less than 3 frames of video!
It was all just dumb luck of course... but serendipity favors the prepared mind, and it turns out you can make *lots* of preparations if you've got nothing else to do while waiting for more data to come in. Menaechmus spent his idle time describing useless ellipses rather than perfect circles; Kepler spent his reading old books while waiting for his colleagues to report back on their telescope studies; and the two made a powerful team despite how stupid and ridiculous it all would have sounded at the time. "Oh, you 'theorize' the orbits of the planets *aren't* perfect circles, but instead the useless shape described by some dead guy? What's next, you *theorize* my dick into your mouth?"
This is hardly the only example, of course. Others include:
• Negative numbers (theorized about *way* before they were accepted as a useful and accurate description of anything in reality)
• Imaginary numbers (same)
• Topology (was the butt of jokes for generations; in other news, the 2016 Nobel Prize in Physics was for Topological Physics and its applications in quantum mechanics)
• Number Theory (G.H. Hardy, a British mathematician, felt so bad about this he wrote "A Mathematician's Apology" about why it was worth doing pure math like Number Theory when it was so obviously useless. His most famous defence was that, since it was so useless, at least it was pure of heart and untouched by war. That was written in 1940. Within a few years, Number Theory turned out to be key to cracking the Enigma, and cryptography more generally.
His other famous example was Special Relativity, of the "mass to energy" kind -- to him, Einstein's work was clearly useless for anything military like bombs.)
• Fast Fourier Transforms/FFTs (invented by Gauss, who thought it was useless; modern applications include detecting nuclear bomb tests through seismic waves, the MP3 file format, and indeed basically anything involving waves)
But the biggest example here would have to be... General Relativity itself. Or at least the mathematics behind it, for describing a Non-Euclidean universe of warped spacetime. By all sense and logic, Einstein should have been starting from zero when trying to describe how the universe isn't actually Euclidean, because who would waste their time investigating this possibility before him when there was precisely *zero* evidence of a Non-Euclidean universe, and *zero* evidence of how it would work? Bolyai, Lobachevski, and Riemann, that's who. They invented tools like Hyperbolic Geometry, Manifolds, the Riemann Metric, Curvature Tensors, and more, so Einstein didn't have to.
Why?
Because they enjoyed it, more or less. Their brains were too active to just be satisfied with the reality in front of them. Turns out they were actually describing a deeper reality all along, even if they didn't know it and were just shitposting at the time. But sometimes, throwing shit at the wall does pay off -- in this case, in *spades*.
So to summarize... that cute story about watching aliens, is actually just summarizing history in a digestable way. What do smart people do when they have nothing to do? They shitpost. They shitpost in such prodigious quantities, about the shape sundial shadows make or whatever, that sometimes they hit gold just by pure dumb luck. But that wasn't even the point, they simply can't help themselves. The allure of shitposting is just too strong. (And hey, sometimes it *does* pay off, 'specially when you've got nothing but time). Why should AI be any different?
Im not *only* talking about yudkowsky, if anything rationalists maybe a bit better at being fishes aware of water
> [humans are] "hooked up to a webcam"
> already happened in history
Your embodied in the world and some guy did just need to drop heavy objects to discover the greeks were wrong about heavy objects falling faster.
I just disagree that you only a meat computer encased in bone sphere, your your entire body; I understand thats gospel here and in the general society. But this is a new idea thats jumping to conclusions while ignoring data such as the possibly that you can treat depression with fecal transplants.
I have a blog post suggesting something even more radical and there michael levins work; but to simplify I believe we are seeing glimpses of a non-reductionist biology and we will *finally* discard to notion that any system evolution creates only does 1 thing at a time with cleanly defined boarders.
> no one will take Heliocentrism seriously if it's simply worse at the job.
People did and I believe I wouldve; heliocenterism was worse at predicting the data for quite a while and pushed by a socail outcast
> Imaginary numbers
> Negative numbers
> but serendipity favors the prepared mind, and it turns out you can make *lots* of preparations if you've got nothing else to do while waiting for more data to come in.
Thats not how I understand that history, and I view this as part of nonreductionism and would suggest the next step is surreal numbers; given a formal system with a unsolvable problem, label the simplest error case as a new primitive, extend the system.
A different way to describe negative numbers is letting the real world rewrite the axoims; animals may have a primitive counting circuit and old languages had no grammar to express negative numbers. Debt is a force of nature that humans took a long time getting a handle on.
"this statement is false" just say "tralse"(logical positivists spent decades trying to argue that such statements shouldn't even exist, before being proven wrong)
Theres the music theory thing of fifths being unresovable.
etc.
Formal systems will not survive contact with reality, this is simply an established fact; label the edge cases as new primitives and see what happens. This is not what the current culture promotes: negatives, imaginary, surreal numbers, follow this pattern but were done slowly and by rare individuals breaking rules.
Id advocate for doing the same with sleeping beauty problem and seeing what comes out of finding some new primitive for Bayes formula.
I think we're talking past each other, because it sounds like... you're agreeing with the story? That people can imagine things they have no evidence for, whether that be General Relativity or imaginary numbers or whatever. Which means they can stockpile ideas like tools in a toolbox, or keys on a key ring, and use them to very quickly open new and unexpected doors/solve never-before-seen challenges, when they *do* finally get the evidence, instead of having to slowly brute force them down. You're... agreeing with the story, not disagreeing.
I.e. the theme of the story can be summed up as something like, "The necessity of empirical evidence, and how to be efficient with it when you don't have a lot." The thrust of what you're saying seems to be "The necessity of empirical evidence, and how you're stuck if you don't have enough of it". Those... are 2 different ways of saying almost the same thing? I don’t see what the disagreement is. Both are ways of saying, "Empirical data is extremely valuable" -- it's just that one describes at so valuable you should be efficient with it, while the other describes it as so valuable you should get more of it. Those are 2 messages that work together, not fight, right?
It's worth being precise about claims that we're roasting. The original claim was that
"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a hypothesis—perhaps not the dominant hypothesis, compared to Newtonian mechanics, but still a hypothesis under direct consideration—by the time it had seen the third frame of a falling apple. "
This is basically just a claim about bit lengths. The entire text of Einstein's "RELATIVITY THE SPECIAL AND GENERAL THEORY" is 200 kilobytes, so that's a _very_ loose upper bound on the size of the turing machine that implements it. 3 frames of video is 3 * 3 * 640 * 480 * 2 = 5 megabytes. AIXI is definitely looking in to that. AIXI directly considers a remarkable number of hypotheses.
> It's worth being precise about claims that we're roasting.
Its not really even claims, its the implicit assumptions, such that precision is highly epistemological valuable
> The entire text of Einstein's "RELATIVITY THE SPECIAL AND GENERAL THEORY" is 200 kilobytes, so that's a _very_ loose upper bound on the size of the turing machine that implements it.
Strong disagree, you must consider the decoder program when talking about thoerical compression(the wikipedia compression contest for example has this rule), so add in several physics text books, cannon english, and human dna minium; im also inclined to believe in bloodline memory and the microbiome being relevant.
The message of "I need help" is not 9 bits long when communicated with an alien, "sos" is *heavily* reliant on precomputed context.
Lets operationalize this, what is the shortest turing machine that simulates the merger of three black holes? Does that seem like a reasonable way to phrase it, and if so are you actually confident that I couldn't produce such a turing machine under 200 kb in an afternoon of work by bodging together stuff I find on the internet? I did a brief try before getting too out of line posting wild claims, and I got as far as running two merging black holes in ~70 kb of C and so it's going to come down to how efficiently once can do C code to Turing Machine transpilation
I dont understand what that program would look like, or what its suppose to prove given blackholes are where the math breaks down and in theory "destory infomation"
The origins of the term are obscure. There is no known etymology of the modern term bodger that refers to skilled woodworkers. It first appears c. 1910,[14] and only applied to a few dozen turners around High Wycombe, Buckinghamshire. The Oxford English Dictionary Supplement of 1972 has two definitions for bodger, one is a local dialect word from Buckinghamshire, for chair leg turner. The other is Australian slang for bad workmanship.[1] The etymology of the bodger and botcher (poor workmanship) are well recorded from Shakespeare onwards, and now the two terms are synonymous
New to my vocab. Thanks Hastings.
Actually I had to look up transpilation too.
I once converted some C++ code for a router for a nuclear power plant control system (a specialized network router) to functionally equivalent C# code.
>Im an anti-reductionist; I believe logical positivism was proven wrong a century ago with gobel incompleteness
What's the relationship supposed to be? LP. Isn't the only support for reductionism. Did you mean Gödel ?
You can tell how well you are doing by how well you can predict : If you can predict most things you are either missing information, or It is not important.
> What's the relationship supposed to be? LP. Isn't the only support for reductionism.
The logical positivists seem to be the extremists in "truth is devised from gathering correct axioms and building up". Which is to say "truth is math" with more steps. Poppers falsification is a weak compromise.
I think darwin would be discarded under the modern bureaucratic science, Begging for mechanisms he never did provide (or couldve, biology is dense with chaotic systems) and when Mendel provided *partial* answers, he may have correctly predicted they were partial answers(tho we will never know, he never commented on Mendels work)
> You can tell how well you are doing by how well you can predict : If you can predict most things you are either missing information, or It is not important.
Its an unknown unknown if you assign an accurate upper bound on predictive power given no access to competitor gamblers.
Given a human market on guessing quantum events, perhaps we all tie for centuries on the current math; maybe some aliens have found an experiment that exposes a lower level and they know a 5 minute experiment to show us the key to the lower level. Or maybe we are correct and reductionism is very helpful for physics, I still dont think reductionism is correct for biology, people move on after getting partial answers "whats the purpose of breast milk" and attempted to make artificial milk before even learning about the immune system, much less our current bafflement about rapid directed evolution by an unconscious system but we do know immunity passes from mother to child.
>It is currently believed in this community that ai will deduce physics of the real world from video: https://www.youtube.com/watch?v=fVN_5xsMDdg why? How do you discover dna without a microscope? How do you discover quantum physics without the double silt?
Agreed (albeit quantum mechanics started from blackbody radiation and the hydrogen spectrum). We build instruments and run experiments for a _reason_.
>Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
I think that's a mischaracterization of what the story is actually saying. I wrote a comment replying to Monkyyyhl here (https://www.astralcodexten.com/p/open-thread-356/comment/78027661), but in short, I think it's describing shitposting, of the kind familiar to 4Chan. Not gigabrained "We don't *need* data" 4D chess moves -- just garden variety "wild ass guesses", and the surprising ways that can actually pay off, just by pure dumb luck. After all, that's what already happened in history! Including with the story of General Relativity itself. (It’s just that this is an analogy the story didn't realize would help explain things.)
>Not gigabrained "We don't _need_ data" 4D chess moves -- just garden variety "wild ass guesses", and the surprising ways that can actually pay off, just by pure dumb luck.
Many Thanks! Ok, I would characterize some of those examples slightly differently.
>Einstein should have been starting from zero when trying to describe how the universe isn't actually Euclidean, because who would waste their time investigating this possibility before him when there was precisely zero evidence of a Non-Euclidean universe, and zero evidence of how it would work? Bolyai, Lobachevski, and Riemann, that's who. They invented tools like Hyperbolic Geometry, Manifolds, the Riemann Metric, Curvature Tensors, and more, so Einstein didn't have to.
Another way of looking at Non-Euclidean geometry is that mathematicians _routinely_ look at mathematical systems and see "Is there a more general version of this? Can we drop one of the axioms and still have interesting proofs?" Group theory can be looked at that way. Topology can be looked at that way.
And it isn't all that uncommon to find that some part of physics, which we'd thought was well described by a very restrictive model, actually deviates from that model - and the generalized model that the mathematicians came up with turns out to _still_ hold, including the newly found deviations from the old model.
In other words, the "unreasonable effectiveness" of mathematics looks a bit less unreasonable when considering that a substantial part of work in mathematics is looking for the answer to "Is there a more general version of this?".
Now, I'm _still_ skeptical about
>"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a _hypothesis..._"
A superintelligence, hooked up to a webcam, would have trouble noticing that the speed of _sound_ is finite, let alone the speed of light, let alone the speed of gravity. I don't think Yudkowsky has really realized just _how_ impoverished the evidence from a webcam looking at an ordinary scene _is_. For most scenes, all of the objects in view won't have more than invisible trace amounts of most of the periodic table. Most scenes will have neither fire nor ice. Most scenes will have no visible chemical reactions. Most scenes won't have enough inertial effects to motivate changing from Aristotelian physics to Newtonian physics.
I'm curious what experiment supremacists believe is happening in physics when Feynman diagrams are invented, or Einstein calls his shot with General Relativity, or we figure out the composition of stars from their spectra. Like, is it that theories can be right be sheer coincidence with very little supporting evidence? Or that they still had lots of evidence? But if it's true that they had a lot of evidence, just interpreted correctly, aren't we right back at where rationalists are?
Like, the world view *as stated* I believe would have said that what Newton did with his laws and orbital mechanics is impossible, that we could have made transistors faster *without* condensed matter physics, or that particle physics during its heyday was impossible.
What am I not understanding about the worldview? Or alternatively, what's wrong with my interpretation of those events?
Many Thanks! I'm confused about what you are asking or saying. Let me try to articulate my perspective on one of the cases you cite.
In the case of deducing composition from spectra, there was indeed a _lot_ of experimental evidence behind this. Lots and lots of observations of spectra of light from materials of known composition - sparks, gas tubes, flames, etc. .
And my main point above shows up here as the need for either a prism or a diffraction grating to see the spectra at all. One cannot just point a camera at a random scene and notice spectral lines. At best, one can e.g. notice the yellow glow of sodium in a flame, but finding the wavelengths of those lines, distinguishing from other nearby lines from other materials, that all requires specialized equipment.
None of this is to denigrate the role of theory. E.g. one needs the theory of energy levels and the transitions between them to turn information about spectral lines into information about the underlying energy levels, and then to tie that to thermodynamic equilibria to understand the temperature dependence of the intensities of the lines - which finally ties it back to stellar composition.
But I'm still confused about whether anything in this conflicts with or confirms your views?
> Lots and lots of observations of spectra of light from materials of known composition - sparks, gas tubes, flames, etc.
Yes, but crucially, none of that experimental evidence was from an actual star. How do you distinguish "you can't discover dna without microscopes" with "you can't deduce the composition of stars without going to one"? Because at the point where you are making statements like "we know burning sodium emits so and so" you *already* needed to have theories based on large amounts of extrapolation about electron ground states and what burning theoretically does. In fact, for something like general relativity I'm pretty sure we had "the evidence" that Einstein used for GR for around 50 years before he came up with it. If experimental evidence is the only thing that matters, why the lag? If reasoning was so important to connecting the evidence to theory, whence anti rationality?
I think by the time you admit "none of this is denigrating the role of theory", it's not clear to me you have anything to defend: someone who thinks they are much better than you at interpreting evidence can just claim that their theory is better, and then what does your philosophy do then? Yet there are strong statements here about how being pro experimental evidence *heavily bounds* the human ability. I just don't see this claim being able to make advance predictions rather than post hoc cope.
>Yes, but crucially, none of that experimental evidence was from an actual star.
True! The best we can do is admittedly a theoretical interpretation of what we've seen from light sources where we _do_ have access to their composition. ( Recently, also with information about the composition of the solar wind. https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/1999JA000358 but that was decades after the original analysis of the solar spectrum and doesn't give us a check on the spectra of any other stars. )
>How do you distinguish "you can't discover dna without microscopes" with "you can't deduce the composition of stars without going to one"?
That's a good point. It's messy. For any question, we always have partial evidence, with varying degrees of indirectness between the evidence and the question we want to answer. Yes, spending effort on thinking through the implications of candidate models is very important.
My only point is that that computational/conceptual effort isn't the _only_ thing that is important. Instrumentation is really, really crucial too. There are many facets of reality which are just too minor under ordinary conditions, too swamped in noise, to see them without setting up the special conditions of e.g. a spectroscope.
>In fact, for something like general relativity I'm pretty sure we had "the evidence" that Einstein used for GR for around 50 years before he came up with it.
For the precession of the perihelion of Mercury, yes, for the deflection of starlight passing close to the Sun (yes, that was a test _after_ Einstein formulated GTR), no.
There are certainly cases where a conceptual snag is the limiting factor. Personally, I find it weird that mathematicians _still_ haven't been able to construct _a_ theory of quantum gravity nearly a century after Dirac extended quantum mechanics to be consistent with special relativity, and more than a century since Einstein formulated GTR. Formulating _a_ theory with the right limiting behavior to match the two existing theories is a conceptual/mathematical problem, not one of instrumentation - though _testing_ it would be an instrumentation problem.
>Yet there are strong statements here about how being pro experimental evidence _heavily bounds_ the human ability. I just don't see this claim being able to make advance predictions rather than post hoc cope.
Well, mostly the limits prevent us from even _seeing_ the phenomena that our instruments can't reach. I'll go out on a limb and say that if there is a sharp resonance in particle collisions at, say, 10^16 eV, nothing in our current theories captures it, because our data is limited to about 10^13 eV, and even the most intelligent extrapolation rarely gives correct answers over three orders of magnitude.
> My only point is that that computational/conceptual effort isn't the _only_ thing that is important.
I think this is sufficiently weak that no rationalist, even a "computational supremacist" like Eliezer, would disagree.
I guess my concern is: what is the *use* of this belief, beyond saying the words "experiment matters". I don't think this thinking pattern allows you to come up with more effective experiments, in fact I don't see how it prevents you from e.g. running more experiments when it's time to understand data, because more experiments = more better. If anything the problem in sciences like nutrition or the social sciences is that people are trying to get p values, rather than figuring out what the fundamental facts of the situation are. Like you *say* that we needed a spectroscope, but in my mind the spectroscope was something more like 80% knowledge of quantum mechanics and its implications and 20% "trial and error". If you know specific details of the history of the spectroscope or other instruments which indicates, for example that people were using it for purposes before they acquired understanding of what they were looking at, then I would love to know your thoughts. (This isn't sarcastic, stuff like the Chinese variolating their families with smallpox, most likely the invention of lots of things like smelting were along those patterns, those just feel like much less common and successful than say, establishing a theory of vaccines or of combustion+ air circulation in furnaces.
> Well, mostly the limits prevent us from even _seeing_ the phenomena that our instruments can't reach.
That seems like a strong argument for theory and against naive experimental fatalism. Particle physics is filled with stories like the discovery of the charm quark: https://en.m.wikipedia.org/wiki/Charm_quark
I *feel* as if this type of called shot would be impossible for an experimentalist supremacy world. How is it not?
See "A Study in Scarlet" and the part about the ideal reasoner:
"Then I picked up a magazine from the table and attempted to while away the time with it, while my companion munched silently at his toast. One of the articles had a pencil mark at the heading, and I naturally began to run my eye through it.
Its somewhat ambitious title was “The Book of Life,” and it attempted to show how much an observant man might learn by an accurate and systematic examination of all that came in his way. It struck me as being a remarkable mixture of shrewdness and of absurdity. The reasoning was close and intense, but the deductions appeared to me to be far-fetched and exaggerated. The writer claimed by a momentary expression, a twitch of a muscle or a glance of an eye, to fathom a man’s inmost thoughts. Deceit, according to him, was an impossibility in the case of one trained to observation and analysis. His conclusions were as infallible as so many propositions of Euclid. So startling would his results appear to the uninitiated that until they learned the processes by which he had arrived at them they might well consider him as a necromancer.
“From a drop of water,” said the writer, “a logician could infer the possibility of an Atlantic or a Niagara without having seen or heard of one or the other. So all life is a great chain, the nature of which is known whenever we are shown a single link of it. Like all other arts, the Science of Deduction and Analysis is one which can only be acquired by long and patient study, nor is life long enough to allow any mortal to attain the highest possible perfection in it. Before turning to those moral and mental aspects of the matter which present the greatest difficulties, let the enquirer begin by mastering more elementary problems. Let him, on meeting a fellow-mortal, learn at a glance to distinguish the history of the man, and the trade or profession to which he belongs. Puerile as such an exercise may seem, it sharpens the faculties of observation, and teaches one where to look and what to look for. By a man’s finger nails, by his coat-sleeve, by his boot, by his trouser knees, by the callosities of his forefinger and thumb, by his expression, by his shirt cuffs —by each of these things a man’s calling is plainly revealed. That all united should fail to enlighten the competent enquirer in any case is almost inconceivable.”
“What ineffable twaddle!” I cried, slapping the magazine down on the table, “I never read such rubbish in my life.”
I don't think AI is that ideal reasoner, but I can see the principle of "by sound logic an ideal reasoner could infer the greater from the lesser" and why that's attractive. I don't accept it myself, but I see the position.
Many Thanks! I could see the position, but I also disagree with it. Even from the examples he cites:
>“From a drop of water,” said the writer, “a logician could infer the possibility of an Atlantic or a Niagara without having seen or heard of one or the other.
Good luck understanding the turbulent flow of a Niagara by observing a quiescent drop...
>By a man’s finger nails, by his coat-sleeve, by his boot, by his trouser knees, by the callosities of his forefinger and thumb, by his expression, by his shirt cuffs —by each of these things a man’s calling is plainly revealed.
Good luck telling one office work job from another by such signs...
Yes I loved reading Sherlock Holmes stories as a kid, but of course had to come to a sad conclusion that one can't tell a story of a man's life just looking at his pocket watch.
Sherlock: “ I deduce that man dressed in rags and begging on the streets is a beggar, that that sooty child on the roof is a chimney sweep, the bewigged older man with a briefcase mounting the steps of the court house is a lawyer, and the well dressed liveried servant helping people on the coach-and-four is a footman.”
Watson: “The guy in the footman’s attire with the Knightsbridge Omnibus Association livery is, you think, a footman?”
Sherlock.” Precisely. My good fellow. Well noted”
Watson: “literally, no shit Sherlock”.
Sherlock. “Despite the vulgarity of your vocabulary, Watson, I take that to mean that you find me free of excrement rather than “full of it”. I take that as a great compliment”.
Would you consider yourself one of the people Im criticizing? Believing physics is math? That we solved biology? Modern science should be all statistics?
Many Thanks! No, I think I'm agreeing with you, and disagreeing with them. I'm particularly disagreeing with Yudkowsky's view that with just video of ordinary scenes and enough computation, all of physics could be rediscovered. For biology - ye gods are there a lot of surprises! Consider all of the drug leads, some from computational studies of binding to target proteins, that fail in clinical trials.
1. COVID wastewater numbers remain low, but SARS2 is still circulating.
2. COVID deaths and ED visits are down, and the weekly COVID mortality rate is lower than at any point since the beginning of the pandemic (of course, mortality will rise during the next wave). In fact, ED visits are still falling ~11% per week. These generally trail WW numbers by a couple of weeks, so I expect the rate of ED visits will level off in the coming week to reflect the WW curve.
3. In the US, XEC's growth rate has plateaued at ~20% of samples and may be falling now. It's unlikely to cause the next wave in its current form (my opinion). For instance, in the UK
4. Over in the UK XEC is greater than 60%, but COVID cases are falling. XEC seems to outcompete the KP.x variants but doesn't increase case numbers. We've seen this pattern before with some other competitive variants (remember Kraken?). My hypothesis is they outcompete the previous vars in a population that is susceptible to those vars, but they don't infect the rest of the population.
5. OTOH, XEC is taking off in Australia. It's at greater than 50 percent in NSW. And a new wave seems to be revving up Down Under.
6. And I need to retract a claim I made in earlier updates that, during interwave gaps, SARS is circulating at higher levels than Flu and RSV by an order of magnitude (per wastewater numbers). After a little research, I learned that Flu A&B shed significantly less RNA out the gut, and Flu RNA is less stable so even though COVID and Flu could be circulating at similar rates in the US population, the WW would show lower levels of Flu. Calculating actual case load from WW numbers is fools game, but they are good for indicating changes in R.
6. I already posted the latest A(H5) avian influenza news in the last open thread.
7. California reported the first Clade 1 MPox case in US. The patient traveled from Africa. It's unclear if it's 1b—which is (supposedly) more transmissible and more virulent. I'm surprised we haven't seen it here sooner if Clade 1 is transmissible as claimed, though.
8. It's started later than normal this year, but Flu season is starting to take off. Biobot seems to show it's more prevalent in the Western US.
9. RSV rates are starting to rise, as well.
10. Mycoplasma pneumoniae infections are increasing. The CDC doesn't officially track these numbers, but Biofire's proprietary Syndromic Trends site suggest that the current rise in cases may not be as dramatic as the MSM has portrayed it.
11. And finally, an actual SARS-CoV-2 recombination event between two vars in one patient and the subsequent transmission to another was documented. We see a lot of SARS2 recombinant (hybrid) vars in the wild, but this is the first time we've seen it as it happened --> https://tinyurl.com/476bkr2w
Thanks! And another reason for careful celebration: it's been four years now, and we still haven't seen any cases of the B Yamagata line of influenza. It might be that the worldwide anti-virus measures have accidentally reduced the number of influenza lines from four to three.
Yes, but influenza showed an interesting pattern. Cases dropped off the cliff to near zero *worldwide* just as the pandemic was revving up in March. I wish Substack would allow us to post graphics in our replies because I could show this better than I can describe it, but within three weeks, Flu cases (from all strains) in the US dropped from 18,000 to 1,200, and test positivity rates dropped from ~25% to ~1%. This same pattern was seen in other Northern Hemisphere countries, too. In a normal flu season, there's a long tail out into April, but the 2019-20 Flu season ended early. In China, where COVID was spreading earlier than the rest of the world, Flu season ended in January (with the same precipitous dropoff in test positivity and cases). Although COVID had been circulating in the US since February, the cases didn't really start ramping up until the 3rd week of March. So, the Flu decline in the US preceded the COVID case surge by about 3 weeks (which is longer than the incubation period of SARS2). Several virologists and epidemiologists posited viral interference as the cause of this pattern, but most experts dismissed the idea. The explanation i saw was that the Flu decline was all due to the farsighted NPIs that the US (and other nations) implemented early. But in the US NPIs weren't really introduced until late April when the first wave was peaking (typo correction: I wrote Alpha, but I meant the original A strain). And many countries never implemented NPIs.
Flu was suppressed worldwide in both the northern and southern hemispheres until the end of 2021, when Omicron hit the scene. The northern hemisphere had a late starting 2021-22 Flu season, but overall Flu cases were low compared to previous Flu seasons. The world returned to normal Flu seasons in the southern hemisphere for their 2022 season, and in the northern hemisphere, Flu came back with a vengeance during our 2022-23 season. Note: China had a steady background of Type B cases all through 2021 that peaked early in 2022. This was when their ZeroCOVID policy was in full force and was the most restrictive NPI policy in the world. Then they had a belated burp of Type A in the spring of 2022.
I don't think the NPI and restricted air travel explanations explain these patterns. I suspect there was some sort of viral interference going on. And when it comes to one pathogen affecting another, we can see a weird pattern between SARS2 and Rhinoviruses. In the US they show a noticeable inverse correlation. When COVID is low, Rhinoviruses pick up. And when COVID is high, Rhinoviruses drop. Or maybe it's when Rhinoviruses are high, COVID is low and vice versa.
Finally, B Yamagata cases dropped after the 2017-18 Flu season. And they didn't really pick up again during the 2018-19 season. I suspect B Yamagata didn't have the numbers going into the COVID pandemic to survive the die-off.
Many have remarked on this seeming paradox. Why does Trump appeal to so many lower-class people despite being a billionaire? It hardly seems paradoxical to the many people in Hollywood whose full-time job is selling an image of luxury to low-class people. The actress partying in her 15 million dollar mansion. The rapper driving his luxury car and bragging about his money. The musician haggling over a multimillion-dollar contract. Of course, the audience needs to be able to relate to the subjects. Such programs focus on the actor, musician, and athlete rather than the derivatives trader because the former are more culturally and psychologically relatable.
Trump has an intuitive understanding of celebrity culture. At the RNC, he didn't invite random people to talk about relatable subjects like not having any money. He invited Hulk Hogan and Kid Rock.
The Democrats' answer was, "Look, we've got a relatable candidate. He's just like you, a middle-aged man who's going bald and wearing a hoodie and has a cat on his lap." And he's "coach." Not the coach of an NFL team; he's the coach of a high school football team. Who trashes J.D. Vance for being a working-class kid who went to Yale instead of Minnesota State University, Mankato:
I think Trump just genuinely likes wrestling. He's attended wrestling events in the past. And I heard he went super into the weeds on MMA (or something?) while on Joe Rogan (I've not watched the podcast episode myself). So Hulk Hogan is not a surprise.
I listened to it, and he did, and he does. Boxing too. (One of the weirdest accusations I continue to hear against Trump is that he's racist, as if to imply the hood-wearing cross-burning sense. Meanwhile, he socialized with black athletes in ways that no hood-wearing cross burner would.)
Low class person: eats McDonalds because it's tasty
PMC striver : doesn't eat McDonald's because that's gross lower class slop
Powerful billionaire: eats McDonalds because it's tasty
The PMC striver class spends their whole time trying to distinguish themselves from the low class yokels, but then demands their loyalty once every four years.
Even during the years my family was homeless and couch surfing, even when I had a 15$ weekly food budget, Mcdonalds was still disgusting overpriced slop.
Innout (or whatever your local innout equivalent is) is right there for delicious fairly priced slop.
If you unironically, un-nostalgically enjoy mcdonalds, you are either a child or have lost all your tastebuds in a tragic lit roadflare licking incident.
Isn't it the other way around? We lose our capacity to taste as we age - it's why kids are super picky while adults are happier with much more strongly tasting foods than kids are.
We have lost our capacity to appreciate how delicious the McDonalds we loved as kids truly is, because we are old now and those taste receptors are all dead and all that remains is the sensation of eating limp ketchup-covered cardboard.
Eh. People say this, but it doesn't work out that way IMO
You don't loose taste until you are pretty old, but some subset of children are hypersensitive to certain tastes to the point where they destroy their ability to enjoy other tastes.
Eg, I could not eat anything with raw onions until my late teens, to the point where once someone put a raw onion on a hamburger bun, remembered I didn't like them, took it off, and I still picked it out. We actually did a blind test afterwards because it seemed ridiculous and I still got it every time.
Eventually this hypersensitivity faded to the point I could appreciate the astringent, sulfur compound taste/sugars and aromatics and such of onions. I wouldn't say my taste has gotten worse as such, because I can till identify a salad that was tossed in a bowl that had a raw onion in it before, it just doesn't bother me anymore.
A smart person puts out a sign saying "pedi-habiliments ambidextrously lubricated and illuminated for the infinitesimal remuneration of ten cents per operation".
A *really* smart person puts out a sign saying "shoe shine: ten cents".
Might be a liberal boomer thing. I think they really did grow up in a beatnik culture that rejected looking rich and successful. Times have changed but the same liberal boomers were picking the candidates.
Your first link is correct. Trump had McDonalds served on Air Force One. That's the kind of 'relatable' that all the spin doctors in all the campaigns can't fake.
And it's not the first time he's done that, and the more stories about being "mocked" for that, the more ordinary people go "yeah, I dunno about him but I sure know *you* are not on my side, so I'm voting for him".
As for the second, Walz emits a slight air of jealousy, never mind that his family circumstances at the start were way better than Vance's family circumstances. And I really don't have much of an image of him or what he is really like, because I don't think the Harris campaign did much with him - he was trotted out now and again, but mostly kept in the background not to take any lustre off the Coconut Queen, for all the good it did. Does Walz eat McDonalds? or another fast food chain? Does he do anything at all, apart from "I used to coach high school football"?
The expression on his face is what sells it. It's clearly not one of the staged "now let's all eat a bacon sandwich to show we're Just Folks", Trump really does like McDonald's and the rest are going along with various levels of enthusiasm.
That's the kind of 'real' that we're talking about in regard to the Harris ads. Did we ever see her eat a Dorito, despite the claim that Doritos are her favourite snack? There was also a photo op with Gretchen Whitmer and the pair of them 'having a beer' together visiting a restaurant in Kalamazoo, but Harris had one sip (granted, you don't want your candidate to get hammered on the campaign trail). At least Whitmer looks like she *would* drink a beer, as she's supping away here:
I will say this for the Harris-Walz... they really do look exactly like Maya Rudolph and Jim Gaffigan. If you wanted to pick a ticket based on the availability of comedians who can play them on SNL, then you could hardly do any better than Harris-Walz.
Trump is terrible for SNL. In fact I've never seen a comedian do a decent impression of Trump, which is incredible.
That part of it I kinda-sorta get: if you start from the premise that the masculinity we've already got is diseased and has to be replaced with something else, "Dad" is certainly a focal-point solution. The unfathomable part is where the campaign got the notion that they could ever appeal to men by starting from that premise.
He was maximum inoffensiveness, which turned out to be maximum ineffectuality. I suppose the lesson to be learned here, from both Hillary and Kamala's attempts, is "if you want to be First Female Ever, do *not* pick as your running mate a white guy named Tim who is a former state governor" 😁
A good test for Trump's 2nd Term: Try to impound funds for SLS (Space Launch System).
SLS, if you don't know, is an abject joke. Multiple billions of $'s per launch. It's been in development for over a decade. It has launched *once*. With Starship on the field, there's really no reason at all to continue funding this to the tune of billions a year. So: impound the funds.
Now you may say: but!...The Congressional Budget and Impoundment Control Act of 1974 blocks the President from impounding funds! Yeah, you're right. Congress, illegally and unconstitutionally, stripped a power from the executive branch that was recognized as a power of the President going back to the very beginning (to stop Nixon, who had an act to grind with the system of absolute Legislative/Administrative branch supremacy). Therefore, with a mandate to restore Government Efficiency and wrestle back power from the deep state, the President should take this opportunity to impound funds for SLS, and restore the balance of power in FedGov somewhat.
And if the Supreme Court tells him no...he should just do it anyway, of course. I'm sure congress will impeach and remove their own (very popular) president for trying to cancel the porkiest of pork pies congress has mandated we fund. No more!
Impoundment is not an enumerated power of the executive branch. It has a long history of being accepted practice, yes, but that's consistent with it merely being an exercise in discretion within the bounds of the laws and appropriations Congress had established up to that point. The history is also somewhat overstated: the practice in the 19th century was rare and was more often deferring expenditure until a later day or spending less than the authorized amount while still carrying out the program. "Policy" impoundments, where the President uses the power to abolish a program or refuse to enact it in the first place, were mainly a thing between the FDR and Nixon administrations. And there is at least one case (Kendall v. United States, 1838) where SCOTUS struck down executive action around spending (payments to mail contractors) on the grounds that the law regulating the amount to be paid was non-discretionary.
Congress does have enumerated powers to appropriate money for various purposes, as well as to make laws "necessary and proper for carrying into Execution the foregoing Powers". And the President has an enumerated duty to "take Care that the Laws be faithfully executed". Taken together, these seem to authorize Congress to restrict and regulate Presidential discretion over how appropriations are implemented.
>"the practice in the 19th century was rare and was more often deferring expenditure until a later day or spending less than the authorized amount while still carrying out the program"
"spending less than the authorized amount while still carrying out the program" is exactly what I'm proposing he do. I'm not saying to cancel the Artemis program, far from it. Just to NOT do it with the pork project launcher that serves *literally no purpose but pork* (Starship, FH and New Glenn can do everything it can do for 1/10th the price).
You can save $2.8 Billion on Artemis 3 by using Starship and not SLS, and you can reassign the NASA employees to do something actually useful. Artemis 3 and Space Science missions can still be executed faithfully, and everyone can keep their jobs. Make Congress defend this boondoggle of a launcher to the public. If they wanna keep it, impound funds and make the Supreme Court stop you. They probably won't, because despite it not being an enumerated power, using impoundments in a manner like this has historical precedent. If they do, *impound the funds anyway* and make congress impeach and remove you if they really, really, wanna keep SLS. Would they? I doubt it, and that would be fine. This is how the system was designed - it's how it's supposed to work. The President not being to fix garbage like SLS and being totally subservient to congress is *not* how the system was designed, and this is like half the reason the Federal Government is such a bloated mess at the moment.
I'm inclined to agree that SLS is a shitty, overpriced program and should be cancelled. If the appropriation had been written for a generic heavy-lift rocket, then I would agree that the President should have the discretionary authority (and the responsibility) to buy the most cost effective rocket that meets the requirements listed in the appropriation. The constitutionally correct time and manner for the President to object to a wasteful or extravagant appropriation is to veto the appropriations bill.
The problem is, I don't think the appropriation is for a generic heavy-lift rocket. If it's specifically for SLS, and it has been properly passed and signed into law, then the President is legally obligated to buy it unless Congress can be persuaded to amend or repeal the appropriation.
As a more general question, you seem to be arguing that the Presidency is much, much weaker relative to Congress than the Farmers had intended. This is a rather novel interpretation of affairs -- I have much more frequently heard the reverse argued.
The President isn't allowed to make decisions on where to spend money, with a few exceptions for emergencies and such. That is a power of Congress: setting budgets and allocating money. The President is the chief executive, to carry out laws. If the President can decide to spend money, or not to spend money, then Congress no longer has that power. This is why Biden wasn't allowed to forgive student loans.
If Congress wants to spend billions of dollars launching things into space, the President must ensure it gets executed. The particulars as to WHERE the money is spent is, indeed, up to the President. But if funding is to be pulled, Congress must do it.
Impoundment is just NOT spending the money congress gave him to spend. Very different from authorizing spending where there was none previously (like the Biden student loan thing).
Congress is allowed the power of the purse, that much is true, but the president deciding "nah" was within his power til Nixon, especially in cases where the President is still executing the law, just more cheaply than expected. Doing Artemis with Starship only fits the bill.
The SLS should be cancelled, even if Starship isn't ready to replace it. But the problem for Trump is the SLS was supported by Republican senators and maintains jobs in Republican states. So he'll need to go against his own party to cancel it.
From what I understand, it was mainly Richard Shelby, who's gone now. Most of the jobs impact (were he to cut the program completely, rather than have them reassigned to something else) would be in Alabama and California. Alabama is red enough he can tell them to pound sand, and the rest of the party is not gonna go against Trump + Elon to save a useless pork project.
Calling SLS a competitor to SpaceX is absurd on its face. It literally only exists because NASA mandated it to a decade ago. It serves no role that Starship or New Glenn couldn't do better / cheaper.
I don't think he gives a damn about what the left thinks of him at this point. The right that's loyal to Trump will probably be fine with it, for privitization/cost-cutting/owning-the-libs reasons.
He has something to offer those red state senators that no other Republican or Democrat can offer: power. Real power. That's all the leverage he needs.
Sure, but power isn't a limited resource even for a President, and every bit of power you trade for one thing is a bit of power you can't trade for something else.
> It's eerily similar to the good times, weak men meme.
But in reverse right? Good times come from generous, cooperative bots (i.e. 'weak'), but eventually everyone is so generous that bad bots take advantage of them, leading to a period of bad times.
When there's a randomized component of the game, where bots might "accidentally" play the opposite of what they intend. TfT can get into an eternal grudge match while always-cooperate recovers smoothly.
I'm not sure if its tested by actual game theorists, but I could imagine one where you take a tiny penalty for each line of code. So tit for tat looses to pure cooperate if those are the only two players.
Penalizing "thinking" (slightly) has crossed my mind as well as a worthy addition to a simulator. TFT has that nice property where you only have to keep one register for the other player. And this feels intuitively like what a lot of people do. (Trump is famously called out here - he seems to only remember the last thing you said about him, and if it was positive, it doesn't matter how much you criticized him before that.)
So, we could incur a small penalty for memory (do I remember what the last move was against me? Do I track each player's last move? Do I track *all* the moves? And the payouts?), and for processing (how much analysis do I do about each move and payout?).
A simulator might have players little smarter than an amoeba, where TFT or WSLS dominate, or smarter players where the dominant strategies start to resemble stock trading algorithms. Or a mix, where the dominant player might implement nearly any strategy at all and dominates anyway because it plays so many games.
tit for tat is a single bit of storage and single operation; tit for N tats is also tiny and can be selected for many rule variants. If a single bit of computation is to expensive, always defect is also computeless and kills always cooperate.
For 1v1 matchups if a single bit is expensive. Always defect beats always cooperate. Tit for Tat beats always defect (the value of the strategy > the expense of computation) and always cooperate beats tit-for-tat.
If we think of humans rather than bots, the time and effort it takes to learn the skills of whose a defector and whose sob story is completely true means that in a world where most people are honest you'll get a lot of cooperate bots.
Where is the appetite for Milei style economic reforms in the US coming from? Both Musk and Ramaswarmy have said America needs such a shock reform followed by a recession and economic pain, but why? Argentina's problems are distinct and not at all like America's. Shock therapy seems to cause unnecessary pain when the problem simply isn't that acute and there are less painful alternatives available. Or do I miss some deep underlying problem that can only be cured by a shock?
It’s class warfare with the signs reversed, that’s all. These policies are hideously unpopular with regular voters of both parties, who want the government to keep its hands off their Medicare/Social Security/farm subsidies/etc.
Cruelty is the point. In Argentina's case, some suffering in likely unavoidable, just like in the Eastern European countries coming out of the Soviet-style socialism. The US doesn't need the same medicine Argentina needs, but both Trump and Musk seem to like hurting people, and Ramaswamy... nevermind, better not flirting with bans, so there.
It's just basic LET'S DO THINGS! style populism. What is being done is unimportant, the important is that it's hard-charging, and drastic, and radical, and that it will fix everything! Wih another set of memes they'd be advocating for the state to DO THINGS! directly.
Musk (re)posting various studies with headlines that sound briefly silly until one thinks more than a few seconds (eg. like https://pbs.twimg.com/media/GcbbzANWkAEofik?format=jpg&name=medium and https://pbs.twimg.com/media/GcbcSDiXUAA83uu?format=jpg&name=medium) rather moves my priors to the direction that what gets done is less Milei-style reforms and more like Golden Fleece style nibbling at the edges. I don't think that drastic Milei-style cuts would be good, but neither would be just turning this exercise into potentially useful research funding being shot down because lol sexbeetles.
The US has some longstanding and well known financial issues that could have been fixed previously but every year gets worse instead of even staying neutral. The national debt increases regularly and the amount of increases keep growing instead of shrinking. We're on track to be paying a trillion dollars a year on interest to the debt next year. Social Security is on an obvious track to insolvency.
There are pretty good reasons why we haven't fixed these issues, but they obviously need to be fixed. It's unsustainable, and everyone knows it. We're unfortunately playing hot potato with these things now, because whichever party actually fixes it is risking not getting elected for a long time, since it is likely to be very unpopular. Milei ran on a platform of doing things to fix long standing problems even though it will cause a lot of pain. If Trump's team can pull that off, then we might dodge the twin bullets of financial ruin and political ruin.
No disagreement here. The question was about the appetite for Milei-style cuts, which Trump is signaling he would be willing to do. I also agree he will find it difficult to pull this off, and will be reluctant to actually try. It will depend significantly on whether Trump is willing to give real power to someone near him (while Milei is the ideologue who believes in what he's doing).
Past practice shows that Trump is unlikely to do that, but I'm trying to be optimistic that things may be changing enough to get it done.
Sure, but cutting taxes is sort of the opposite of reducing the national debt.
The last time the federal government ran a surplus was under Clinton, who worked out a compromise with the GOP to raise taxes and cut spending. Bush inheritied an 100b annual surplus which he blew up with massive tax cuts. Obama worked out a Clinton-like compromise with GOP Speaker Bohner but the House GOP rank and file were dead set against giving Obama anything, so it died before even coming up for a vote. Now we are going to have even bigger deficits. I don't see any possibility of a fix without both tax increases and spending cuts.
>Clinton, who worked out a compromise with the GOP to raise taxes and cut spending.
Clinton and Congressional Democrats raised taxes a lot and cut spending a little in 1993, with zero Republican votes in Congress. Then Congressional Republicans pushed large spending cuts and small tax cuts in 1995-1996 against Clinton's objections. Clinton and Congressional Republicans did compromise on something that still cut both spending and taxes, but by less than Republicans had wanted. That compromise was projected to balance the budget by 2002, but windfall tax revenues from the Dot Com boom lead to the surplus appearing early, in the 1998 fiscal year.
The combined effect was a combination of tax increases and spending cuts, as the tax cuts (even before Clinton pushed back and got them reduced) were a lot smaller than the 1993 tax increases, but it's important to note came in two stages with different Congresses approving each stage.
>Bush inheritied an 100b annual surplus which he blew up with massive tax cuts.
Also massive spending increases, both domestic spending (mostly Medicare and Education) and war. And the Dot Com boom ended. We could have afforded any two of the tax cuts, the domestic spending increases, and the wars and still had a surplus again by the time the recession ended, but not all three at the same time.
IIRC, Clinton also cut the deficit by borrowing from the Social Security fund in a manner that didn't come due until the Bush administration. (Imagine if Gore had won and had this dropped in his lap.)
You don’t recall correctly. You are probably remembering somebody arguing that the unified budget numbers were misleading. The argument is reasonable, which is why the Social Security program is “off budget” in the first place, but is as applicable to the Bush years as the Clinton years. There was nothing to “drop into the lap” of a hypothetical Gore Administration.
The “on budget” numbers exclude “off budget” programs, including Social Security. The “on budget” numbers don’t distinguish between debt held by the Social Security Trust Fund an debt held by the public. For that reason, it is not possible to cut the “on budget” deficit by borrowing from Social Security
The “unified budget” includes all Federal programs, including “off budget” ones like Social Security. During the Reagan Administration, Congress passed a plan prosed by the Greenspan Commission that caused the Social Security Trust Fund to run a surplus in order to build up funds to pay for the baby boom retirees. This resulted in the unified budget showing a much smaller deficit or larger surplus than the on budget numbers. In fiscal year 2000, for example, the unified budget showed a surplus of $236 billion whereas the on budget numbers showed a surplus of only $86 billion, mostly due to Social Security running a surplus. This discrepancy increased during the Bush Administration; in fiscal 2008 the on budget deficit was $642 billion but the unified budget had a deficit of only $459 billion. So the unified budget deficit was reduced due to borrowing from Social Security, but Clinton had nothing to do with that.
I don't quite understand what you mean by "on budget". Looking it up, it appears to mean what I took to be the plain reading, which is spending within prior limits, as opposed to unforeseen spending. If so, SS ought to be on budget, not off, and you say it is. So that's probably not what you mean. The meaning I found seems closer to what you're calling "unified budget".
If I just move on to your last paragraph (where most of the meat is anyway), I have trouble tracking down the legislation you cite. I see the "Social Security Amendments of 1983", aka H.R.1900 (https://www.govtrack.us/congress/bills/98/hr1900/text), but my skim of it doesn't tell me anything obvious. ("Surplus" doesn't appear anywhere in the text, so if it happens, it's implied, either by raising FICA or lowering payouts, and indeed, a lot of the bill does that, so, maybe?) I can see how your account explains an apparent reduction of deficit / increase in surplus. It's possible the explanation I heard was referring to HR1900 (or whatever the proposal became), but I can't tell for sure.
If I happen to run into someone who argues that Clinton merely borrowed from SS, though, I can pass this along and see if it jibes with their evidence.
I do notice, meanwhile, that if this plan happened for the reasons you describe, then it wasn't general fiscal prudence on Clinton's part, but rather a specific concern for the Boomer retirement surge that went back to Reagan, which is new to me.
>I don't see any possibility of a fix without both tax increases and spending cuts.
And the former is not going to happen. It is unlikely that the Republicans will even permit the 2017 tax cuts to expire as scheduled at the end of 2025.
I have ended up following more US politics than local politics without even trying. Does anyone have a good writer to follow for succint, rational takes on the state of European/German politics?
Very simplified but: Germany is having economic issues. The German government wanted to spend more (mostly domestically) and did its equivalent of hitting a debt ceiling (in this case a debt brake law). Scholz fired his finance minister after the finance minister's party didn't agree to either increasing taxes or changing the debt brake. This led to his coalition no longer having a majority in the legislature. So he called an election to sort it out.
Overall it looks like the Center Right is going to win right now. Though they will need to form a coalition and with whom is an interesting question. But we'll have to wait for the vote totals which we should get in February.
> Say there's a magic button that, when you press it, reduces your LDL cholesterol by 30%.
I really hope this is market research, because having a probiotic you could take that would reduce LDL by 30% would be awesome.
If I were trying to sell it, I wouldn't even want to get it FDA approved, I'd just mass produce Cholester-yogurt or whatever and hedge with a bunch of "this MAY reduce cholesterol, as part of a complete breakfast" or whatever, like the oatmeal people do, and let people do their own research and make their own conclusions.
But then, I've always been a fan of "go high volume and small margin" as a strategy versus the usual drug / bio strategy of "niche and high end."
You've hit the nail on the head, my guy! The good lawyers at General Mills have spent decades carving out the Cheerios Exemption, AKA Structure/Function Claims.
Probiotics are foods/supplements in the US, so the primary regulatory burden for a new food is just proving safety as long as you don't make disease claims. Cholesterol is really in the sweet spot here because everyone knows that "number go down" means lower risk of heart attack and stroke.
And yes, I'm hoping to make this as widely available as possible, but the logistics involved in anaerobes-as-probiotics mean substantial running costs and geographically limited rollout
I hope you'll publicize it on your Substack whenever it's ready. I'd probably pay $1k, and I've got no history of triglyceride problems or CVD in the family.
No idea. Not nothing because I have a. Just been reading a book by a woman who made herself tetraplegic falling off a horse and b. Learned that spinal column strokes can have the same effect. But the question isn't answerable unless you quantify the changes in absolute and relative risk caused by pushing your button. In the most favourable case perhaps £10000? I am oldish and have high LDL for which I get free statins from socialised medicine.
My copay after insurance for the lowest possible dose of rosuvastatin is $6 a month. My LDL wasn’t really that high but the statin reduced it more than 30%.
I had a physical a couple months ago and doc said my cholesterol was high. By about 25%. Since then I’ve been trying to eat right. I eat walnuts and fucking mackerel all day. I’m losing it.
Statins are a no go because I’m religious about only taking drugs that get me high.
If someone said they could perform surgery to do what your button did, I’d probably get it done if my plan covered it. If all I had to do was press your button, no side effects, and witness medical magic, $4855 seems right, but not a penny more.
Walnuts and mackerel *every* day? I like mackerel but even that is a bit much 😁
God alone knows what are cholesterol reducing foods, I've eaten nearly everything suggested and nothing worked (niacin helped a bit, but that was the "your face starts burning like a cherub after taking it" version and I couldn't tolerate that long-term).
Right, I'm saying you, personally, right now. Given your risk of heart disease and your disposable income.
And yes, statins exist, but plenty of people on statins still have heart attacks or strokes. I suppose, if you take statins now and wouldn't have to after pressing the button, you could calculate a floor value based on how much you'd expect to pay for statins over the course of your lifetime—but risk of cardiac events is basically a function of the integral of LDL with respect to time, so even someone who's still gonna have to take statins should theoretically derive a few QALY of benefit from the button.
Sure but statins are a thing you have to keep taking.
One of the things I dread about getting older is the idea of becoming a Person Who Takes Medication. Once you enter that category there's no getting out of it, it's just an ever increasing daily pill load until you're dead. I'd definitely pay the $4855 to put off needing to enter that category.
This is an oddly specific medical conditionality, both the condition and the amount.
It seems to me the benefit to this would allow you to eat anything you want without the problems associated with higher LDL cholesterol. As such, it equates to how much you're missing from not eating those things. In my case, it might be worth something like $100 a year.
I don't really think I ought to disclose my medical conditions and medications in this format, so it's tough for you to see my context.
My LDL is fine. My total cholesterol is fine. But my triglycerides are way too high and I can't shift them. So I'd pay a reasonable amount for that kind of reduction.
I want a number! No wrong answer, but half the point of this question is to figure out how accurately people value their lives, given that heart disease is still the #1 cause of death.
I could say it would be worth €1 million, or €10 million, or €100 million to me. I don't have anything approaching that kind of money, so it's a useless question.
If you mean "out of the money you have right now, as a once-off payment, how much?", then I'll pay €100 as that is literally the maximum I can afford. EDIT: This does not mean I only value my life at €100, I mean that's all the spare money I can give for something not immediately urgent like the house burning down. See what I mean about stupid questions?
Now, if you're finished asking dumb questions and getting dumb answers, why not ask people "would you take this? would you value the opportunity?" instead of this mania for "put a number on it! it don't mean nothin' if you don't put a number on it!" for the rationalists here. If you want numbers, heck, let me throw the US interest of $1 trillion dollars (as quoted in a different comment) on it as the kind of meaningless dumb valuation of "put a number on it" - yes, I'd very much like to get my triglyceride levels down since the actions and medication I've been taking to date haven't helped. If a magic button fixed that, I would be happy to press the magic button.
>This does not mean I only value my life at €100, I mean that's all the spare money I can give for something not immediately urgent like the house burning down.
No, see, this is helpful. Like, everyone values their life more than their possessions, but that's the thing about uncertainty and probability—you never know that you should've prioritized "not having a stroke" higher until you're in rehab trying to re-learn how to talk. I expect if you were somebody who'd already had his first heart attack, you might be willing, for instance, to sell your phone, or take out a small loan so you could pay €200 or €300 to hit the button.
FYI, I'm not just doing thought experiments here, I'm doing market research. Genuinely trying to figure out how to price a one-time intervention with potential to persistently lower cholesterol. No idea if it works for triglycerides yet, but if you want I'll keep you posted.
Ahhhh. Slightly more than comfortably full. Tryptophan addled. House a on the warm side of comfortable.Time to lie on the sofa in front of a football game I have no intention of paying attention to and uh uhhhnn….
A hypothetical situation/question that has bugged me since childhood: A ball is thrown at a wall. The distance between the ball and the wall can be measured and expressed mathematically. As the ball approaches the wall, that measurement becomes a smaller and smaller number. Numbers are infinite, and numbers can be always be divided and made smaller (in my understanding). My question - as the ball nears the wall, the numerically expressed distance can always be made into a smaller number - how does the ball, numerically speaking, ever actually make contact with the wall? This may be a dumb question, but hey, if you don't ask... Thanks in advance.
There’s a collision detection system in the simulator, it works pretty well.
These are not trivial problems at all. If you suppose that the ball moves at constant speed of 1m/s toward the wall, and starts out 3m away, then the distance to the wall is (3-t)m and as the time t approaches 3, the distance goes to 0. There's nothing special happening at t=3 from a mathematical sense, though physically the ball contacts the wall at that point and you need to refine the model to go any further.
Actually, Newton developed calculus to deal with problems of the type you're describing. The way calculus is taught in schools and universities obscures this fact, but the basics of the subject are all about making sense of processes which happen in continuous time, like the one you describe.
You'll read this and think I haven't answered the question. Actually setting calculus on a completely rigorous mathematical footing took around 250 years, and the effort of many mathematicians. (Real numbers are already surprisingly difficult to describe formally - and then you're looking at processes involving these to describe the real world.) To get a completely satisfying answer to the question, you would probably need to go through a sequence of undergrad level calculus courses - but hopefully this gives a sense of what's going on.
Keep in mind that time works the same way; every time you remeasure the distance from the ball to the wall, you remeasure the time it takes to travel there. Eventually you hit... is it Planck, or Heisenberg? Either way, there's a distance at which things are no longer things, and then the ball changes direction.
Technically, the ball never touches the wall. The ball is made of atoms, the wall is made of atoms; if you put the atoms too close, they repel. At a tiny but nonzero distance the force becomes strong enough to make the ball bounce.
(If Chuck Norris throws a ball at a wall, a microscopic black hole is created at the place of collision, and then it gets complicated. EDIT: The atoms would probably just fit in places between other atoms and the ball would get stuck inside the wall, possibly leading to an explosion.)
You're looking for solutions for Xeno's paradoxes, which in the modern era broadly get solved with calculus. I'm not super up on the exact solutions, but it's something to do with limits?
EDIT: Here's the actual solution, explained for people that haven't thought about math formally for a long time: https://sciencehouse.wordpress.com/2020/12/06/math-solved-zenos-paradox/
Gebru and Torres, two dissident researchers who became famous as coauthors of the renowned "Stochastic Parrots" article, recently published a new paper critical of the rationality sphere and EA:
https://firstmonday.org/ojs/index.php/fm/article/view/13636/11599
They coin the awesome-sounding acronym TESCREAL to describe the sphere, and then argue that it has roots in the eugenics tradition. Are you aware of this article, and have you considered responding to it?
> dissident researchers
In what way are they "dissidents"? They're representatives of the boring, narrow hegemony that rules intellectual life, trying to extend that hegemony into one of the few corners they don't control.
Once I strip aside the authors' tone of moral panic about the followers of TESCREAL beliefs (Transhumanism, Extopianism, Singularitarianism, Cosmism, Rationalism, EAism, and Longtermism) their arguments boil down to (a) these are all influenced by eugenics-based ideas, and (b) and most of their AGI hopes and dreams are not based on sound scientific nor engineering principles.
As for the first, they admit that the dream of breeding/creating superior humans goes back long before the early 20th-century Eugenics movement, and they fail to tie the current eugenical beliefs to those of the previous century. But they try to tar the current movements with old stains of the old movement.
And as for the latter, I tend to agree with their statements. A lot of the TESCREAL beliefs are attractive to smart people who think they know more than they do (and Charlie Stross has noted how much of their beliefs have been influenced by old science fiction tropes). I see the TESCREAL belief systems as being mostly harmless crackpottery. Of course, certain billionaire followers of TESCREAL beliefs could do a lot of social and political damage if they're given the power to do so.
Reading the article, commenting as I go, I am not promising that I will make it to the end:
> we have seen little discussion of why AGI is considered desirable by many in the field of AI, and whether this is a goal that should be pursued
Apparently no one reads Yudkowsky these days.
> In this paper, we ask: What ideologies are driving the race to attempt to build AGI?
In this comment, I answer: it's about money and power; the things that transcend all ideologies.
> the acronym “TESCREAL” denotes “transhumanism, Extropianism, singularitarianism, (modern) cosmism, Rationalism, Effective Altruism, and longtermism”
At the first sight, this sounds like a complicated way to say "nerds". (Later we will see how much the words can actually be substituted, but if yes, that would kinda reduce the entire finding to: "the people trying to build AGI are... nerds". Which is true, but not really surprising.)
> The idea of eugenics can be traced back to the origins of the Western intellectual tradition
Yeah, I guess if you want to talk about transhumanists and singularitarians, you need to start with the recent examples, such as Plato and Aristotle.
> second-wave eugenics arose in response to new technological possibilities associated with genetic engineering and biotechnology. Such technologies opened the door to human “improvements”
When the authors say "eugenics", they basically include anything that concerns human DNA. If you worry about your child potentially getting Tay–Sachs disease, you know who else worried about people having inferior genes, right?
The link between eugenics and transhumanism seems to be that eugenicists wanted to improve humans by improving the genetic pool, and transhumanists want to achieve the same goal by abandoning the biological limitations completely. Different methods, the same goal.
Extropianists were the first organized modern transhumanist group. Singularitarians are also transhumanists. This... apparently makes the belief that an artificial intelligence could be smarter than humans an eugenicist belief, by the power of association. (I wonder whether a belief that Google can find things online better than you is also a fundamentally eugenicist belief. Or that a calculator can do multiplication faster than you.)
Never heard about cosmism before, so whatever; I find it plausible that it is also somehow related to thinking about the future and technology (which, as we have already established, are fundamentally eugenicist topics). Rationalists... also believe that AI is a big thing, and that it would better for the poweful AI to be good rather than evil. (Those assholes.) Effective altruists want to alleviate poverty... I am sure there is something sinister behind that, too, let's wait for the punchline... ah, yeah, some of them also worry about the future humans. And longtermists, those by definition worry about the future. (Remember, worrying about the future is an eugenicist thing.)
> Indeed, transhumanism, Extropianism, singularitarianism, and cosmism are examples of second-wave eugenics, since all endorse the use of emerging technologies to radically “enhance” humanity and create a new “posthuman” species.
OK, still in 20% of the article, but seems like this is the crux. If you accept the premise that trying to improve humans in any way (such as curing hereditary diseases or augmenting humans by technology) is basically the same thing that Nazis tried to do, only using modern technology, then... either you become a Nazi sympathizer, or you need to ban all efforts to cure genetic diseases or augment humans. Anything else means being in denial; and luckily we have two smart professors here to expose your hypocrisy.
> The TESCREAL bundle shares certain “eschatological” (relating to “last things”) convictions. As with religions like Christianity, these take two forms: utopian and apocalyptic, which are inextricably bound up together.
The belief in either horrible or glorious future is also in Marxism, just saying. On the other hand, effective altruists are mostly worried about the dark present. (I think that rationalists are also quite concerned about people being stupid right now.)
> The reason for concern is that emerging technologies are expected to be (a) extremely powerful; (b) increasingly accessible to both state and nonstate actors; and (c) dual-use, as exemplified by CRISPR-Cas9, which could enable us to cure diseases but also synthesize designer pathogens unleashing an “engineered pandemic”.
Here I would like to ask the authors whether *they* agree with this assessment factually, or not. Because if they do, then it also puts them in the category of people worrying about the future. (And if they don't, then what's the point of obsessing whether the technologies are connected to the sinister TESCREALs?)
> in 1996, Yudkowsky expressed concerns about superintelligence, writing: “Superintelligent robots = Aryans, humans = Jews.
Ah, Yudkowsky being Yudkowsky. But what exactly are the authors trying to say? Are they suggesting that Yudkowsky proposes the extermination of Jews/humans as a desirable thing to do? Or are they just taking things out of context in order to make their point?
> “Intelligence,” typically understood as the property measured by IQ tests, matters greatly because of its instrumental value
Uhm... yes?
> The obsession with IQ can be traced back to first-wave eugenicists, who used IQ tests to identify “defectives” and the “feeble-minded.”
Ah, I see the logic. In a parallel universe where Nazis sent *short* people to death camps, it is probably a taboo to talk about human height, and if you suggest that putting things on the top shelf can make them difficult to reach for some people, you get cancelled.
(PART 2)
> Table 1: “TESCREAL bundle” of ideologies.
A list of people and organizations to get cancelled.
> Effective accelerationists believe that the probability of a bad outcome due to AGI is very low
Yeah, that's pretty much the same thing as Yudkowsky says, except with "not" at the beginning.
(Would that make the acronym TESCREALEACC?)
> how TESCREAList groups are steering the field of AI toward the goal of creating AGI.
This feels like giving too much credit. Are they saying that without these people, it wouldn't occur to anyone to build an LLM?
> In 1955, four white men officially launched the field of AI
Ah, the field is forever tainted by the association with crackers. :(
> many researchers in fields currently associated with AI, such as natural language processing (NLP), machine learning (ML), and computer vision (CV), explicitly distanced themselves from the term “AI,” in part because it became associated with unfulfilled grandiose promises. Nonetheless, some groups continued to work toward “artificial general intelligence,”
So, both the former and the latter agree that the AIs we have now are not AGIs? And that is somehow controversial? Or why the framing of "some researchers work on X, but others [sinister music starts playing] work on Y"? Isn't it quite normal that different researchers work on different things? It seems like the authors are trying to hint at some dark controversy here, but to me it sounds about as sinister as saying "some biologists study mice, but others [sinister music starts playing] study frogs". Like, yeah, that's how research works. (Actually, I may be wrong here. That's how research works in natural sciences. In some other fields, it is probably very important that the researchers agree with each other and all do the same thing, otherwise they get cancelled?)
> How, then, would researchers know that they have achieved their goals of building AGI? They need to know how to define and measure “general intelligence.” Unsurprisingly, these definitions rest on notions of “intelligence” that depend on IQ and other racist concepts
(facepalm)
> For these reasons, Keira Havens, who has written extensively on race science, asks those attempting to build AGI: “Why are you relying on eugenic definitions, eugenic concepts, eugenic thinking to inform your work? Why [...] do you want to enshrine these static and limited ways of thinking about humanity and intelligence?”
Mussolini promised to make trains run on time. Why are you still using trains?
> One conjecture is that the resulting AGI will be so intelligent that it will figure out what the best thing to do is in any potential situation.
I'm confused. So did they read Yudkowsky or didn't they? Or were they just looking for a quote to take out of context, and ignored everything else, including the part where he *literally* called exactly this thing "the sheer folly of callow youth"?
https://www.readthesequences.com/The-Sheer-Folly-Of-Callow-Youth
> TESCREALists have been able to divert resources toward trying to build AGI and stopping their version of an apocalypse in the far future, while dissuading the public from scrutinizing the actual harms that they cause in their attempts to build AGI.
Uhm... sorry about mentioning this Yudkowsky guy all the time, but you included the movement he started in your acronym... and you pretend to be experts on what the movement is about, so... maybe it would be nice to get familiar at least with the very basics?
> As an example, one of us worked as a hardware engineer designing audio circuitry for devices such as laptops. Some of the tests that we performed as part of our work included drop testing, constantly dropping devices to understand the manner in which their functionality degrades when they are exposed to shocks, placing the devices in extremely cold or hot environments, frequently restarting them [...] we argue that the first steps in making any AI system safe are attempting to build well-scoped and well-defined systems like those described as “narrow AI,” rather than a machine that can supposedly do any task under any circumstance.
OK, I really don't want to put your expertise in restarting laptops in doubt. But, you know, these AI things sometimes turn out to work differently than you would expect. For example, I think that scientists in 1960s believed that we will make artificial intelligences understand human knowledge one topic at a time, by building expert systems and cleverly designing the semantical nets. They would probably be shocked to hear that the state-of-art AIs of 2024 are good at poetry and painting, but suck at math. And yet, here we are. The line between the narrow AIs and AGIs may turn out to be similarly unpredictable; it can happen much later or much sooner that we expect, and may be caused by an improvement that we will not suspect of having exactly that impact.
(THE END)
OK, my summary would be that the authors identified a group of people whom I would describe as "nerds thinking about the future, specifically the technology", and correctly noticed that there is an overlap between various groups.
The connection to eugenics is mostly made through "trying to improve things" and "believing that intelligence is a thing". I understand that both of these things are *problematic* for some people, but I also think those people are... how to put it politely... not thinking clearly, and probably not even interested in thinking clearly.
Some of their arguments require *a lot* of cherry-picking. If you fail to notice that the effective altruists mostly care about reducing the suffering that exists *today*... or if you believe that Yudkowsky wants to build the AGI as soon as possible, and believes that by the virtue of being generally intelligent it will get all the answers right... then I strongly doubt your expertise on the individual elements of the "TESCREAL" boogeyperson you invented. It's more like you sometimes need to make things the very opposite of what they are, in order to make your point.
Congratulation on finding your powerful and rich enemy responsible for manipulating the society and science towards their sinister goals. You know who else did that, right?
> If you fail to notice that the effective altruists mostly care about reducing the suffering that exists *today*
To be fair, is that even the case anymore? I thought that disagreement about that was what caused Open Philanthropy to splinter off of Givewell.
Yes. As usual, it is 10% of reality that gets 90% of the attention, but see e.g. here: https://www.givewell.org/charities/top-charities -- malaria, malaria, vitamin A, some unspecified vaccines. Okay, I guess the vaccines are about reducing suffering in the future, but it is still different from what the "TEASCREAL" model suggests.
Maybe somewhere lower on the list there is a longtermist project, and that is what makes some people freak out. Or maybe there are also separate lists by categories, and the longtermist category contains longtermist projects. But there is a difference between "there are some X somewhere among Y" and "Y is about X".
On the other hand, top 2 (and also 5th and 6th) careers recommended by 80KH (as of today) are AI risk-related.
https://80000hours.org/career-reviews/
I coin a slightly less awesome acronym HGT, which means Hitler, Gebru, and Torres. I claim that HGT has roots in Nazi traditions, and I have historical data to prove that at least one third of HGT was actively involved. That sounds quite damning, in my opinion.
Long story short, you can choose a perspective that supports your conclusions. As a rationalist, you should also reflect on how much your conclusion reflects the territory and how much it merely reflects the perspective you chose. To get the best results, you should aim to "carve reality at its joints", i.e. let yourself be guided by how things actually are, rather than your arbitrary definitions, no matter how clever they may sound (they may or may not be correct). For example, I am pretty sure you can find a connection between "TESCREAL" and eugenics tradition, just like you can find a connection between the eugenics tradition and progressive politics. The question is, is the former connection stronger than the latter? (If not, why do you focus on it so much? Instead of e.g. writing an article about "TESCREAL" having roots in the progressive politics?)
...okay, now I am going to actually read the article.
EDIT: One of the names sounded vaguely familiar, took me a moment to remember why:
https://forum.effectivealtruism.org/posts/yAHcPNZzx35i25xML/emile-p-torres-s-history-of-dishonesty-and-harassment
Torres wrote a response, if you want to go down that rabbit hole:
https://mileptorres.substack.com/p/effective-altruism-is-a-dangerous
Still haven't read the article, but now it seems to me that the alleged main connection between eugenics and "TESCREAL" is... consequentialist thinking, and worrying about the future of humanity. Technically true. (So it's like, someone says: "I would prefer if AI didn't kill all humans", and the response is: "so what you're saying is that you are worried about preserving the future of the while race?")
Well I skimmed through it... and the paper doesn't seem that wrong at the object level. Hell, even the last survey Scott posted did show that the readership for the blog is relatively pro-eugenics (but only when he didn't use the word "eugenics"). It's just that the authors' values are completely idiotic. What is with these people's obsession with identity politics and "marginalized communities"? Calling people "racist" just for trying to improve humanity is just... well, it's certainly proving their opposition's point that all of this is necessary.
>What is with these people's obsession with identity politics and "marginalized communities"?
Buzzwords and such part of an institutional jargon designed to advance that institution's interests or satisfy certain needs.
I agree. Unfortunately, these values are dominant in Western academia and their supremacy is jealously guarded. Some high-profile EA should probably make a response, otherwise I'm worried that good EA ideas will see even more opposition from the academy.
Well that shouldn't matter, academia is going to get ideologically purged soon anyways. Though, I guess we'd have bigger things to worry about at that point...
By whom?
By the new administration. Do you think the Republican party sacrificed all of their values just so they could maintain the status quo? The people, and the people that lead them, want change, and the only way to accomplish that is by rooting out the rot at the top. Not just in academia, but in corporations as well.
Of course, they almost certainly hate EA as well for being filled with leftist vegans, so uh... Good luck with that.
The Trump administration last time around couldn't even manage to purge the Trump administration. This time around I think they'll be ever so slightly more effective, but I doubt they'll manage a purge of academia.
Is the rationalist movement privileged? Ok, stupid question. I'd say the answer is an obvious yes. What I mean specifically is privileged in the economic sense. I.e. something that some wealthy people can afford to do/be but less wealthy people can't afford to be/do. And when I say rationalist I mean someone who esteems truth for truth's sake. Feel free to argue for or against this first question, but I take it as a given. It seems obvious to me that some people are getting subsidized by their jobs/career to be rationalists while most other people get subsidized to shut their mouths and do what they are told. So my actual question is, in what ways can one incorporate rationalist principles in their life if honing these skills may actually end up hurting them economically? Do you see it sort of like a Socratic matter of principles (I'll be broke as hell but at least I'll know I was being honest with myself), is the world really just dog-eat-dog and poor people should just excise rationalist thinking from their lives if they know what is best for them? Other ways of looking at this? I ask because I work as a secondary ed teacher and I've found that there are aspects of my job where being a rationalist is helpful/good but I think it is valuable to bifurcate rationalist thinking (truth for truth's sake) and just play the political game a lot/most of the time.
Good question. Why not take a data centred approach?
How do we measure privilege? It's going to depend on the society you're immersed in, and your own values. But most people would agree that wealth, education, a 'good' job, social standing and various freedoms/civil rights are a reasonable measure of this. The first three are fairly easy to measure. In the amount of time I'm willing to spend on this, I can compare the ACX survey to US demographic data.
Something like 85% of Scott's readers have a third level degree, upwards of 20% have a doctorate. Less than 40% of the population in the US has a bachelors degree. Approx. 2% have a PhD and at most 5% have doctoral level qualifications. So readers of the blog do have educational privilege. Since the ability to do a PhD requires years of forgone earnings, it's reasonable to infer that blog readers are coming from a family background which is higher socio-economic status.
Unfortunately, the survey data just reports numbers for income with the majority of responses censored. You could ask Scott for the data if you wanted to do more analysis, but I think it's likely that average and median earnings are well above the US average. Participants did self-identify their SES but I would treat this with caution, it's better to collect 'hard' data rather than self-descriptions, particularly when something like 'upper-middle class' is context and country dependent. Personally, I think the level of education and amount of free time required to read blogs like this and post in them means that they're biased to the privileged.
Recently, I've been at odds with my manager over small issues. E.g. they wanted to rearrange the schedule of a meeting I was chairing so they could arrive and deliver the welcome address 30 minutes in - I insisted the welcome be at the beginning, they could speak on other topics if arriving at the end of the meeting - they rescheduled to arrive on time. I don't think this is in any way a rationalist approach - to me it's common sense. I had the background and sufficient social standing to make my case persuasively. I didn't play the political game though - and I'm lucky to have a boss who is (mostly) willing to listen to pushback. I'd be curious to hear if a card carrying rationalist would have approached this differently - to me it was a mixture of common sense and my own stubbornness.
Not especially; the non privileged rationalist don't call themselves rationalists because that term is intensely weighed down (unfairly) in the consensus by your Theilites and your SBFs and your weird cryptofascist orbiters.
Eg, they decided not to pursue lots of money for moral reasons; they are out there digging wells or being teachers or doing the type of research that doesn't end in you being super rich for inventing the Juicero.
I would say that it's hard not to be a rationalist if you are at all introspective.
A) The Big Yud often said that "rationality is the art of winning". I.e. winning is the goal, rationality is a tool. If rationality precludes you from winning, discard rationality and win. He typically relies on Zen motifs to get his point across. C.f. "The Twelfth Virtue" [0] and "Something to Protect" [1].
B) The wealthy are privileged ipso facto. Dependency (of any kind, including financial dependency) represents exogenous influence. Unfortunately, specialization has made modern society quite interdependent. So it often is, infact, in your best interest to adeptly navigate Baudrillard's Simulacrum [2]. Another way of thinking about money: money is "slack" [3], and slack is options. Including the option to be intellectually-honest.
> So my actual question is, in what ways can one incorporate rationalist principles in their life if honing these skills may actually end up hurting them economically?
C) To paraphrase The-Bugman-Who-Must-Not-Be-Named: "you have to live like you're in enemy territory". If the powers-that-be don't like what you're teaching your students, you need to realize that you're basically distributing samizdat.
[0] https://www.lesswrong.com/tag/twelfth-virtue-the
[1] https://www.lesswrong.com/posts/SGR4GxFK7KmW7ckCB/something-to-protect/
[2] https://www.lesswrong.com/tag/simulacrum-levels
[3] https://thezvi.substack.com/p/slack
These are helpful! Thanks
Big Yud also said that it's about truth, and assumed the formulations were equivalent , and they're not.
Hm... I don't remember anything like that. Got links? Not gonna lie, it's been a long time since I read The Sequences. And I never completed the entire corpus. So I'm prepared to be surprised.
Whatever the case, I stand by my initial comment, insofar as winning is ultimately more important than truth. (No, this isn't a carte blanche to act like a pathological liar. But Kant's radical honesty, for example, is a kinda bonkers imho.)
> Is the rationalist movement privileged?
Movements are not privileged; people are. Some rationalists are rich, some are poor, most of them probably somewhere in the middle. Or slightly above the average, considering their typical professions.
> something that some wealthy people can afford to do/be but less wealthy people can't afford to be/do.
My personal freedom of speech is mostly a result of living in Eastern Europe where no one gives a fuck about what you think. If I lived in America, I would probably watch my tongue more carefully. I am *not* wealthy enough to not need a job, which dramatically limits the things I can do (and in America would also limit the things I can say).
> It seems obvious to me that some people are getting subsidized by their jobs/career to be rationalists while most other people get subsidized to shut their mouths and do what they are told.
Yeah, your type of job determines what kind of idiots you are surrounded by.
> Other ways of looking at this?
It seems like some good (for self-preservation) habits from communism are also useful in the woke era, so here is one:
* Don't think.
* If you think, then don't speak.
* If you think and speak, then don't write.
* If you think, speak and write, then don't sign.
* If you think, speak, write and sign, then don't be surprised.
Personally, I would think, and only talk to friends or talk pseudonymously on the web.
> I work as a secondary ed teacher and I've found that there are aspects of my job where being a rationalist is helpful/good but I think it is valuable to bifurcate rationalist thinking (truth for truth's sake) and just play the political game a lot/most of the time.
Yeah, you can use some rationality at your work, without saying explicitly that this is what you are doing. (When someone important says "there is a dragon in a garage", say "of course there is", and then continue using the garage as usual. Which is what many people actually do.)
Thanks Viliam!
Others may chime in with different takes, but, for me, rationalism is about forming confidence estimates related to beliefs, it is not about saintly purity in how you choose to act or what you choose to say.
Taking an example from inside your question: imagine a situation where you are called to make a statement and you see two choices (a) what you believe to be true, (b) what you believe to be politically expedient. Rationalist methods could help you in the following ways:
(i) identify a false dichotomy and determine a third (or even more than one) alternative thing to say/way to act
(ii) form and update an estimate of the probability that (a) is a better/worse choice than (b)
(iii) phrase your belief (a) in a way that does not have the assumed political cost (because you indicate openness to other possibilities, you support your statement with evidence, you shade your statement probabalistically)
This is really helpful. Thanks Joshua!
Ted Gioia just dropped this on his substack - https://www.honest-broker.com/p/strange-and-dazzling-things-are-happening
tldr - NY Times reports, paywalled, (https://www.nytimes.com/2024/11/18/business/media/substack-politics-newsletters.htm) that Musk tried to buy Substack, but failed. Ted is a big enough whale on Substack that he has met Substack CEO Chris Best in person. Ted's comments are worth quoting, so here they are -
Last year I found myself across the table from Substack CEO Chris Best—and I hoped he would satisfy my curiosity on one matter. So I asked:
“Chris, I have a question. Has Elon Musk made an offer to buy Substack?”
Then I added:
“Appropriate answers to this question are (1) Yes, (2) No, or (3) I prefer not to say.”
Chris didn’t answer immediately. He thought it over for a moment, then responded: “I prefer not to say.”
In this soft-spoken way, Mr. Best showed his wisdom. Although he works hard to empower journalists, he’s smart enough to know to watch his words around those pesky folks. We’re as leaky as sieves, and only half as reliable.
But the New York Times reported yesterday that Elon Musk did attempt to buy Substack last year. Musk also hinted that he would merge it into Twitter and let Best run the combined companies.
The Times also reports that Best rejected this offer.
And he didn’t even fret about it. It was a “short-lived discussion,” according to the Times’ three sources.
That proves that Chris not only has wisdom, but also possesses integrity and strong core values. When he says that he supports indie writers, he really means it.
If Musk’s plan had been accepted, Chris would have enjoyed a huge payday in the transaction, and also might have found himself running the most powerful media force in the world right now.
But he didn’t hesitate to walk away from that.
I want to commend Chris and the entire Substack team. They not only talk the talk, but walk the walk—which is rare in the media world right now. Or in any world right now.
The article makes it seem so obvious, but...
First, saying "I prefer not to say" is not necessarily the same as kinda-deniable "yes". Maybe he rejected Musk. Maybe he is waiting and hoping that Musk will offer more money. Maybe someone else offered money, and he is trying to make the other person think that they are competing against Musk. Maybe he even said yes to Musk, but they have some kind of deal that the website will officially change hands on January 1st 2025, and until then there is an embargo on this information.
From business perspective, the Substack model seems great. By taking a fixed fraction of subscriptions, the financial interests seem aligned; Substack makes money when the authors do. The possible downside is that this can make the platform attractive to all kinds of conspiracy theorists and scammers (I imagine that writing Flat Earth News could be quite profitable), but if you advertise yourself as a platform where no one gets censored, that already goes with the territory.
But technically, using Substack can be quite frustrating experience, and frankly it seems like it's no one's priority to improve it. It took years just to make ACX comment section usable, and it still could be improved a lot. Or the bug where if you replied to a comment in an e-mail, your reply got inserted incorrectly in the comment tree. (That one is probably fixed, but that also took years.) I haven't written a post lately, but I remember it to be also frustrating. Like, how am I supposed to write articles containing code examples, if all existing character style options are so huge that you can barely fit 40 characters in a line? Or why do I have to choose whether a word is monospace or a hyperlink, but it cannot be both? It's obvious that the intended type of article is "type a lot of text, drag and drop a few pictures and a YouTube video, insert five 'subscribe' buttons, and call it a day", which I guess is where most money comes from. Yeah, it makes sense economically, but that means that if your technical skills are even slightly above zero, Substack will constantly make you cry.
So, maybe don't sell to Musk, but at least hire someone who is technically competent and who cares.
>That proves that Chris not only has wisdom
What makes you confident of that? Maybe accepting the offer would have been good for Substack. Why are you so sure that it wasn't?
"Cannot confirm or deny" is usually the best response to questions about possible takeover offers even if they haven't occurred.
> "Cannot confirm or deny"
It's also the preferred statement from the CIA, which is known as the Glomar response.[1]
[1]: https://en.wikipedia.org/wiki/Glomar_response
Well, at the very least we can be sure that the negotiations didn't go well:
> This month, Mr. McKenzie publicly called Mr. Musk a propagandist “with more conflicts of interest than El Chapo.”
Are we sure Musk didn't just lowball him? There's no practical reason for him to overpay for Substack like he did with Twitter, considering that this site is already pretty right-leaning.
And reading the article more closesly, Substack still isn't doing very well financially... Musk has more than just money as leverage, given that he immediately retaliated by hiding posts with substack links, not to mention his new political connections. He'll get what he wants eventually.
Not that I think that'll be a massive issue anyways; all that'll happen is that everyone at Substack gets fired and the site gets merged into X. And also he'll probably stop renewing most of the contracts with the writers here. I guess that'll suck for Scott.
It would suck not just for Scott but also all of us here in the comments.
In what way? Musk has shown a commitment to free speech that few others in media have.
Until it's speech that he doesn't like. Then, um, not so much
One of the weird things happening in the amazing nowadays is the number of people eager to jump to defend Musk. Like, you know he's a man with immense wealth and power, and is quite capable to looking after himself. He's also not running for any popularity contest. So... why do so many spend so much effort defending whatever he does? Jumping to counter every lightest slight in his direction, like he was a precious baby or something...
That's weird, because I got the opposite impression, that recently it is incredibly popular to accuse Musk of... well, nothing specific, just being generally the worst. Seems to be so since the moment he bought Twitter, but maybe I got the timing wrong.
I don't even like the guy; I think he is neither a saint nor a devil, just a smart rich guy trying to become even more rich, who seems to spend too much time online shitposting.
Is there somewhere a list of his "crimes" that I could read?
For the same reason that lots of people jump to criticize him for no real reason. He's become a symbol of the culture wars, so everything he does gets framed that way.
Also I don't think it's fair to characterize my comment as "jumping to counter every lightest slight". I genuinely don't think him buying substack would be bad for anyone and don't understand why someone would reflexively assume that it would be.
Oh? What's an example of speech he's suppressed purely because of its content?
He makes rules on the fly and interprets hazy policies in a way that justifies him banning, suspending or pushing voices he doesn't like off the platform. When he took over I was curious to see what twitter with free speech would look like, but unfortunately these are not the actions of someone with a genuine commitment to free speech.
https://en.wikipedia.org/wiki/December_2022_Twitter_suspensions
https://www.vanityfair.com/news/2023/04/elon-musk-twitter-still-banning-journalists
https://www.theguardian.com/commentisfree/2024/jan/15/elon-musk-hypocrite-free-speech
Doesn't substack have competitors by now?
Wait! Wait! The Webb Telescope disproved the existence of Dark Matter?! I guess I must have been snoozing. The massive galaxies that Webb observed in the early universe are an excellent fit for MOND, and their existence that early in the universe pretty thoroughly falsifies the Dark Matter models of galaxy formation.
Stupid question: if Modified Newtonian Dynamics (MOND) is a better description of gravity, how does this affect the value of the Cosmological Constant? Does Dark Energy also go away? I don't know enough about MOND because I never took it seriously. I guess another Internet rabbit hole awaits my exploration.
https://iopscience.iop.org/article/10.3847/1538-4357/ad834d/meta
Hmm, well I've been following Mond for a while now. Tracy McGaugh's blog is full of stuff. https://tritonstation.com/
Later edit;
I just want to add that Mond is less of a 'theory' and more of a 'fit to the data'. If gravity at accelerations below a_0 became a 1/r force law, (rather than 1/r^2) then that fits all the data.. up to some point. Mond fails at the mass level of galaxy clusters, it doesn't fit the third hump in the CMB data. nor the bullet cluster... There's something missing from Mond.... My favorite idea is some hot dark matter... sterile neutrinos. (I mean first of all I love the name.) Oh and that Mond is true... whatever mond is... beyond a fit to the data.
I don't think it can be concluded yet that DM or that the age of the universe has been falsified nor that has MOND been proven.
On this topic, the empirical evidence for MOND seems better but that isn't a 100% exclusion for DM.
DM can explain current star formation history, if and only if models of star/galaxy formation "innovate".
Be it a non constant IMF, super eddington accretions, assumptions about black holes and dust prevalences and properties, etc.
We are at the limit where nearly 100% of baryons needs to be converted into stars which sounds highly unlikely though. There has been some pre JWST era models that match most current observations. Moreover there are many possible more exotic catalysis mechanisms of galaxy and star formation, among which some might require new physics or standard physics but in practice non standard physics (few people will understand what I mean here), e.g. via dust based Mock gravity, rydberg matter or Zeldovich radiation.
Most importantly the debate should not be dumbed down to MOND or DM, one both can coexist, two there are many extensions (interacting DM, etc), three both can be false, the true theory might be quantized inertia, etc
My personal belief is that the age of the universe is at least moderately wrong.
The MOND existence will be forever settled via GAIA DR4 in 2026 (is settled in 2025 but ESA will gatekeep the data because they do not follow open data science)
BTW here is the biggest issue in LCDM by far, and you'll never hear of it, like in most sciences, the most interesting discoveries are absolutely unknown because well there is only one person on this planet reading the scientific literature extensively (me)
https://arxiv.org/abs/1706.04771
Oh Mond is by no means complete.. it fails at higher mass levels, galaxy clusters and above.
In reference to your paper, how would a missing dust signature in the CMB support the existence of dark matter? Would the presumed distribution of dark matter—clumped around galaxies and diffuse to a near-perfect vacuum in intergalactic space—reduce the density of dust in intergalactic space?
Also the Z>14.3 barrier has already been broken, the information is just not public yet
Uhm, could we please restore the ancient tradition of Open Threads without politics? I am not really interested in reading about Trump every day.
Yeah, I agree. The only problem is where (or who) draws the line?
Banning the word "Trump" would alone get us 80% there. Adding "woke", "Israel", and "Palestine" would move it to 95%.
This is Scott's substack, and he has the power and the right to draw whatever lines he wishes. But in the past, Scott used to post a separate political open thread every week (month?) that allowed political discussions and arguments. He did a good job snuffing out the political crap on the other threads—but I suspect it became too onerous for him. And so, I'm forced to wade through political crap to get to any interesting philosophical or scientific discussions. This is one of the reasons I haven't re-upped my subscription to this sub.
Are those not the hidden open threads for the paypigs?
No. Plenty of Trump stuff there too.
Yes please.
I second that motion!
https://apnews.com/article/gaetz-trump-fbi-justice-department-248b46ba0c882dd46d661568e8bd3bd7?utm_source=firefox-newtab-en-us
Matt Gaetz has already withdrawn from the Attorney General nomination, having previously resigned from the House, meaning he's currently not in any political offices. This removal of Matt Gaetz from Congress has the chance to be the first win of the new Trump cabinet, although I don't know if his resignation will actually hold this early in.
It aeeems that nobody who writes about these things in the news and elsewhere is familiar with negotiating or what they call in VEEP "cock-thumb" it was immediately obvious to me that that's what trump was doing with Gaetz.
Bernie Sanders' latest doomed crusade is limiting American arms shipments to Israel: https://www.jpost.com/american-politics/article-829168
Put aside the question of whether Sanders is right or wrong at the object level - my opinion is that Israel has a right to self-defense, crushing Hamas/Hezbollah is in the interests of all non-insane people, and while all conflicts have occasional outrages the IDF has overall done a good job in a challenging urban environment - the war is divisive in America. Whether those opposed are right or wrong isn't the question. Everyone has a right to their opinion and proportional say in how American power / tax money are used, dumb though their opinion may be.
What's a rough percentage of Americans opposed to this or some policy generally that you feel is sufficient opposition to be placated? Does it need to be a majority? 50%, 25%, 10%, etc.?
Would it be smart politically for the Democrats to cater to that opposition? For Trump? Would it be unifying or equally divisive for the government to say that the cost in domestic dispute is too high and Israel should source weapons elsewhere?
I do not know if I have good answers.
Democrats basically alternate between cheerleaders for the military (who quickly fall out of favor when wars go badly) and diehard pacifists (who get credibly accused of being America-hating foreign agents).
It's smart to have both wings so that a candidate from either side can be put forward when the situation calls for it. In previous conflicts, the pacifists have been either really, really right (Iraq the second time, e.g.) or really, really wrong (Iraq the first time, e.g.), so it helps to have a wide range of stances.
Republicans do the same - Donald Trump represents his party's isolationist wing, whereas George W. Bush (the previous Republican president) represents the interventionist wing.
Maybe both parties support wars they like, and oppose ones they dont.
Hillary Clinton (2016 nominee) voted for the Iraq War. Barack Obama (2008 President) voted against it.
Meanwhile, Barack Obama campaigned on the idea that Afghanistan was the "good war". His Vice President, Joe Biden (2020 President) withdrew from Afghanistan, letting it fall to the Taliban.
So the Democrats liked the Iraq War enough to vote for it, and to vote against it. They liked the Afghanistan War enough to campaign on winning it, and then to lose it.
I think it's easier to describe that as two separate ideological wings living uncomfortably with each other, rather than one group exercising a consistent ideology.
Wasn't Biden left with a fait accompli by Trump who'd made continued occupation untenable by the time he'd left office (by making a deal with the Taliban, letting the Taliban infiltrate governmental zones, and making physical preparations for total withdrawal)?
That's one way to frame it. Of course, when he was running for President, he discussed withdrawing from Afghanistan - look at the 2020 Democratic Party Platform and you can read their plans of Afghanistan under the heading "Ending Forever Wars". It's pretty much identical to Nixon's peace with honor that ended the Vietnam War.
Contrast that with 2008 - the chapter on Afghanistan is called "Win in Afghanistan", with sentences like "We will send at least two additional combat brigades to Afghanistan, and use this commitment to seek greater contributions–with fewer restrictions–from our NATO allies."
So I find it hard to believe that Biden wanted to surge additional troops into Afghanistan, a la Obama, but was thwarted by Trump. I think it's easier to believe that both Biden and Trump wanted to end the war in Afghanistan, but blame the other party for the inevitable bad press.
There's 13 years between the 2008 campaign and the 2021 withdrawal; plenty of time to change your mind based on the "facts on the ground" (a rare actual military use of the cliche). Also, two different people.
I'm trying to argue that the Democrats have two different wings - one pacificist/isolationist, and the other interventionist. I'm comparing the most recent Democratic President with the previous Democratic President - those two are unfortunately separated by 12 years. That's an unfortunate, but unavoidable part of the data set I'm working with.
You could however look at a single slice in time, with a single chamber. Look at the 2003 vote on the Iraq War - 81 Democrats voted in favor, 126 voted against it. For comparison, 215 Republicans voted for it and 6 against. The Republicans at the time were very much interventionists. The Democrats? They were split - not perfectly split, but a 60-40 split is a very real split.
Sounds about right as long as "ones they like" are just "ones they started".
I'd love to just poke my head briefly into the Gore-invaded-Iraq timeline to see how it's playing out.
And sometimes, they get confused about which wars they're supposed to support and which ones they aren't supposed to support.
if you're gonna traffic in atrocity denial at least give us the courtesy of not calling your critics insane
one party in this conflict is producing refugee camps and mass graves and it's not Hamas or Hezbollah
That's a fair point on my wording. I apologize. It was a quickly written post and I shouldn't have used that framing. I was thinking more of Hamas in light of Oct. 7th but it was much too forceful and broad. Consider it withdrawn.
If you punch me in the face, don't complain when I rain punches on you. "But you're the one punching now! You ought to stop!"
Israel's objective is not to produce refugee camps or mass graves, but to prevent being attacked again. Were they trying to wipe out the Palestinians, they're doing a bad job given their resources. They are showing as much restraint as reasonable in limiting civilian casualties.
No it isn't. Remove 1 terrorist organisation and another pops up.
Israel is doing exactly what I said they would do in the immediate aftermath of October 7th.
I wrote about it here 2 days after the massacre:
https://ydydy.substack.com/p/a-point
And I reviewed that prediction a year later here:
https://youtu.be/OR67EQNvc1k
And last week I published the actual solution.
https://ydydy.substack.com/p/hasbara-vs-hasbara
I don't expect much support on account of society being composed of shortsighted stay-the-course cowards who reserve their strongest antipathies for outsiders and who will die defending their leadership absolutely no matter how bad they are. I assume this comment wil go over like a like a lead balloon thus proving my point.
Even the people in Room 101 loathed Emmanual Goldstein.
O how far from Eden we are.
The second paragraph here is a good point.
The first paragraph not so much. It's perfectly possible for a response to be excessive. If someone punches me once, and then I "rain punches down on him," continuing to just beat him until well after he's been turned into an unconscious broken piece of meat, I'm well past the point of a reasonable self-defense and should expect others to intervene, and possibly to be convicted of a crime. A lesser crime than I would be if I was unprovoked, but "he hit me so anything goes" is not the relevant standard. At a certain point, when "you're the one punching now," you really *ought* to stop.
Hamas is still punching, though, so the whole analogy seems misguided. It's not like Hamas surrendered and Israel just kept bombing them.
> I'm well past the point of a reasonable self-defense and should expect others to intervene, and possibly to be convicted of a crime
I'm sure you can get away with that if you have the blessing of the US government and military. The justice system exists to serve a need that isn't applicable in military conflicts, or any other state-sanctioned violence for that matter.
Well, it wasn't a perfect analogy, as none is. But you're right from one standpoint: one punch doesn't give license to punch forever. My point was more that you can't just call to stop just because you're losing.
And Israel, in this case, is trying to prevent the "first" punch from happening again. I do believe that if they were convinced it would stop, they would cease hostilities. But they think, justifiably, that if they stop it will just give their enemy a breather to prepare for another devastating attack. And then what is the proper response to prevent future attacks? Genocide?
what is the proper response to a conflict that was started by their forefathers but then they turned it up to 11 since ... let's say 1995?
they tried to ignore the simple fact that their co-tenants are not happy with the arrangement they forced and continue to force on them.
of course - turns out! would you believe it? - that after decades of marginalization said co-tenant's behavior shows, rather concerningly, heightened aggression.
and as a complete surprise, the hardliners continued promises that they can keep the peace ... failed.
....
yes, this is a big fucking Situation (waaay beyond Troubles level), with heavy military aspects, so the use of military force is not unfortunately some kind of "crossed line".
the moral arguments usually focus on the cruelty (and in some sense futility) of what the IDF does, and how it does it, not that it has "the right"
not to mention that - as far as I understand - the concept of self-defense is not well defined or really applicable for states occupying territories
so, on one hand, the IDF could go slower, be more careful, the tunnels won't run away after all. the only semi-sane objective is long-term occupation, otherwise the tunnels will get built again
on the other hand this is just more of the same, this guarantees more instances of mistreatment. (and even typing the words international peacekeeping mission led a laugh track fading in from somewhere.)
... and all that said, what is the proper response to being in this fucking terrible pickle? well, a bit more brotherly love would go a long way, but apparently there's been some mixup in the warehouse and thus recently folks are mostly getting war, far-right authoritarians and mayhem in other shape and form.
Most likely it has to start with showing self-restraint. (While not neglecting this elusive self-defense either.) But as long as Israel cannot really commit to dealing with its own fundies it will have an ample supply of counter-fundies :(
Agreed on both counts. Though when to stop is a more complex consideration than if you happen to be the one happening to be punching at the moment.
I'd expand a little on the second paragraph. The delta in mass graves / refugee camps between Israel and Hamas don't - as far as I can tell based on Hamas' stated goals and actions - reflect a difference in intents, only capabilities.
Ark is leaving out the part where he got punched in the face a while before the current punch... and punched again a while before that... and before that... and pretty much on a regular basis going back several decades, including a few rains of punches here or there.
In other words, the current punch isn't the only punch; it is the latest, and delivered as if there will be more, and maybe even a rain if permitted. That's why he rains punches back.
It's even worse than that. A few punches ago, he punched back more than once and "you're the one punching now" *was* the argument used to get him to stop... and then he got punched again. So that argument fails twice as hard now.
Are they showing restraint in torture?
>the war is divisive in America.
No, it's really not. It's a problem for a bunch of loud yutzes, but they poll in the 5% range, and literally no one else cares except for the people (>10%) who are wondering why Israel hasn't been tough enough.
>Everyone has a right to their opinion and proportional say in how American power / tax money are used
Yeah, no. Literally no one gets a proportional say in how their tax dollars are spent, most especially including the yutzes. I'm not sure why you think this either is or should be the case, but no, that's not happening.
The right answer for Dems is just what they've been doing, which is to pretend the yutzes don't exist and that they don't have to do anything about them.
> Yeah, no. Literally no one gets a proportional say in how their tax dollars are spent, most especially including the yutzes. I'm not sure why you think this either is or should be the case, but no, that's not happening.
You're correct that we, wisely, don't take an Athenian style vote on major decisions or budget line items. However, everyone does get an equal vote when it comes to elections - theoretically equal, state level distortions for the Presidency not withstanding, at a minimum equal at the level of your state.
Also agree with Peter Defeel below. If you're American you have ~335 million brothers and sisters. There are always going to be significant portions that see you as a yutze. You will be a yutze to your future self and vice versa. But there are no correctness requirements for voting. We don't throw out the votes of someone motivated by fear of subterranean lizardmen or equivalent. Like it or not, everyone with the right to vote gets a say even if their reasons are objectively or, to you, deeply stupid.
The question is also more of about the "ought" versus the "is". Yes, an administration with sufficient support from elected officials has the formal power to do whatever it can. But imagine a scenario where Examplestan is in an armed conflict with New Somewhereville. Magically reliably polling shows 50.1% in favor, 49.9% against. To what degree as a matter of national unity ought the views of a minority position be considered?
Well, that's why the two-party system was nice: it made sure the populace didn't get to have a voice in regards to actually important matters. Of course, the Republicans ended up defecting, but a one-party system effectively does the same thing.
*Moldbug sweatily pounds out 10,000 overwrought words on the Sovereign.*
Replete with italicized irony.
A ceasefire is supported by the majority of democrats and the majority of people under 40.
> Literally no one gets a proportional say in how their tax dollars are spent, most especially including the yutzes
Everybody is a yutze to someone else.
Almost everyone who supports a "ceasefire" is imagining a particular set of terms for that ceasefire. Which they imagine to be obvious and achievable, so that as long as the war continues they'll say "stop it, you guys, cease fire now! [on my terms]". If the firing ceases and their terms are not in place, they'll say "that's not what I meant and you know it!". It's win-win for feeling good about yourself and blaming the bad stuff on other people.
So, what are the terms for the cease fire you're implicitly advocating, and what's the line for "that's not what I meant by ceasfire"?
It is impossible for an electorate to have in-depth, technical views about how things should be done. Instead, it is ideal for them to have an understanding of what they want done and for the political leadership to figure out how to reconcile their desire with the hard realities of the matter. The pro-Palestine people want peace and the upholding of international law. The US could work towards these aims by leaning on Israel a lot harder than they have been doing and forcing Israel to restrain itself when it comes to damage done to civilians, journalists, and other noncombatants. Israel could otherwise pursue its ambition to eradicate Hamas under these constraints.
That's awfully vague as a response to "what are the terms of the ceasefire you are advocating". But I notice that, at the end, you've got Israel continuing to try to eradicate Hamas.
In which case, what you are proposing is almost definitionally *not* a cease-fire. Unless you have a serious plan to eradicate Hamas without shooting at them, or unless you weren't serious about Israel continuing to try to eradicate Hamas.
Hamas delenda est; all else is commentary. If somewhere in that commentary is a serious, actionable plan to delenda Hamas with fewer civilian casualties, great, but all I'm seeing is a wish or a command for such a plan to exist.
It would be expedient for police in Europe to carry firearms everywhere they go and to shoot first and ask questions later like their American counterparts, but it has been deemed by democratic political processes undesirable that expedience should take such precedence over other concerns, like the possibility of innocents getting shot. Just because eradicating Hamas would be made easier by ignoring the international community doesn't mean that it is desirable from a broader perspective or an ethical one. America is sacrificing its own values by supporting Israel, as it is going against the very rules-based global order it created, currently manages, and has always championed, where nations are supposed to submit themselves to international consideration rather than act unilaterally out of realpolitik and pure power-consideration. America is torching its own moral authority. Even from a pure realpolitik perspective this seems to be more damaging to America than beneficial.
The practical limits Israel would need to adhere to would be no greater than those America accepts when waging its foreign wars. It is surely still possible, but an effort towards optics would have to be made, a fig leaf adorned.
A quick search of recent polling didn’t give me any answers to these questions but suggests younger Americans are less enthusiastic about military aid to Israel than older ones.
I personally like to see Senators fostering debate on big foreign policy issues and would like to see it happen more often, even if it doesn’t lead to legislation.
Sanders is 83. He ran in the 2016 and 2020 Democratic primaries. Both times, it was clear the Democratic establishment was against him winning and both times Sanders ultimately failed. Any hope he has of one day being US President has to be a massive longshot at this point. People know that he doesn't speak for the Democratic party as a whole.
So, he might as well just voice his conscience, his honest opinions on things.
He never actually joined the Democratic Party, still hasn't. It was odd that a sitting legislator who'd pointedly declined to be a member of a political party for decades, could enter himself into primaries of that party. (Will Rogers' Depression-era joke was "I don't belong to an organized political party -- I'm a Democrat.")
I suppose by letting a non-party member run in their primaries and seeing him get some substantial votes in them, the Dems are then basically stuck with him as far as public impressions go.
Politicians should, as clearly as possible, say ahead of elections what they will do if elected, and in which cases they will exercise their judgement in new situations and based on what.
Bernie sanders is acting predictably. So is Trump. So are most politicians, from the perspective of their voters, on this issue.
Agreed. Voting is an extremely lossy signal though. A strong isolationist, America First style Trump voter may be completely opposed to all military aid but still vote for him although he says that he'll increase support to some countries.
>Would it be smart politically for the Democrats to cater to that opposition?
Yes. I assume Republicans already get the voters who really want to support Israel. Democrats should offer the opposite option for voters who really don't want to support Israel. Not doing so leaves an opening for Republicans to take at least some of those votes. Eg, Trump did a lot of outreach to Arab Americans before election day.
https://www.youtube.com/watch?v=BGJdclFhtO4
Similarly there should be parties that stand for isolationism so voters can express that preference, which is why I think it's a good thing that neocons have been pushed away by the Republicans as the Democrats have increasingly embraced them. I also don't like the chattering classes trying to set the agenda by making certain preferences verboten, whether it's restricting immigration or asylum (increasingly less so today), being against support for Ukraine, etc.
That seems odd. Is there a minimum threshold of votes before a party should take an opposite approach to grab those voters? Pro-murderism seems open, for instance, but seems like it would do more harm than good.
The Republicans absolutely do not get all of the pro-Israel votes, and the Democrats are pretty evenly divided on the issue (with older Democrats overwhelmingly in favor of Israel while younger ones are only somewhat disapproving). That's a lot of voters to lose, and pretty much doom for the Democrats if they did. Supporting Hamas over Israel is one of the biggest losers in electoral politics, similar to things that nobody would consider part of electoral politics (even if not as bad as being pro-murder).
>Is there a minimum threshold of votes before a party should take an opposite approach to grab those voters?
Yeah, it's about 5%. If you try to grab single-issue voters below 5%, you'll end up losing more votes from either people who quietly disagree with that 5%, or who think that anyone who would go after that 5% is a sleazy charlatan. The Ds somehow realized that anyone who backed Hamas would be thought of as either a lunatic or a sleazy charlatan, and miraculously avoided doing so. I know! I'm as surprised as you are!
Where is this magic 5% coming from. This seems to be arguing in bad faith - perhaps the 5% is in favour of the destruction of Israel or pro a one state solution. What Bernie is arguing for is the end to the recent war which is popular with democrats and young people.
I get the feeling you're not really responding in good faith. Obviously some positions are far outside the Overton Window and it's pretty unreasonable for you to ask me to name some specific number of votes that a position should give before a party offers it as an option. "Implement/compel a ceasefire" for example is not outside the bounds of acceptability as you made out. Second, saying that Dems should, say, campaign on decreasing military aid to Israel is not saying that Dems should also campaign on giving that aid to Hamas instead.
I'll admit I'm somewhat incredulous about the support for Hamas that we see. From the perspective of Democrats, though, it seems electorally insane to offer any support for Hamas, and only qualified support for the Gaza Palestinians (who apparently like and/or support Hamas). You offered the perspective that it would be "smart" for Democrats to court that vote, and I pointed out two issues - 1) Support for Hamas is not popular in the US, even among Democrats, and 2) Support for Israel is far more popular, including among Democrats.
It's a losing proposition. I also don't think "chase single issue voters even if their goals are unpopular" makes any sense generally. It's perfectly fine for neither major party nor any minor party to support the KKK. It's also more than okay if no party supports Hamas.
I've never argued against someone pro-Israel on the internet before this, but I finally understand why the other side gets annoyed by online pro-Israel commenters strawmanning them constantly to the point that it honestly looks like the Israeli government is actually paying people (you) to do it.
There are a sizeable number of people who'd like their governments take actions that Israel, or at least the current government, might not like but are not pro Hamas.
Is there a third choice besides "supporting Hamas" and "supporting Israel"?
Here is my peace plan for Israel/Palestine. It may take a few years, but it is still more realistic that any proposed alternative I have seen:
Every Jewish family should be legally required to have at least 10 kids. When those kids get 18 years old, the entire territory of Israel and Palestine becomes one huge democratic state, and everyone who lives there will get full citizen rights.
This is a perfect win/win solution for everyone. The Jews get their Jewish state. The Arabs get to live in peace at the place where their ancestors lived. Everyone gets democracy and human rights.
(Alternatively, make it 20 mandatory kids for every Jewish family, and also conquer Lebanon. You are probably going to do that anyway, so at least let something good come out of it.)
Sure, but for some reason it seems most people aren't really going for any of the many third way options. From my perspective, it seems the anti-Israel/pro-Palestinian side wants to force the issue to be binary in order to juice their gains. Mixed would by far be the most popular, followed by pro-Israel, if all options were on the table. The only way to get more than a lizardman level of support for Hamas is to force the dual choice.
I'm at the Whipple museum of history and philosophy of science at Cambridge university. So cool. Here's info about an instrument built by John Harrison to measure longitude at sea.
https://collections.whipplemuseum.cam.ac.uk/objects/15687/
I've been to a similar museum in Harvard, but this is even smaller and less overwhelming...
Btw, Longitude is a great book for young kids.
Why does a kid even need to attend school if they can go to such a museum with guidance?! Only for a social life I suppose. And in some states because it's a legal requirement.
What, is every city in the world going to have a museum covering every topic?
Great question. No, a museum won't be all of the curriculum, but a learning aid. And dies it have to be in your city? Give the $17k per kid per year to the homeschooling parents to use for such expenses.
The govt needs to be good at some minimal checks on the process and results. The results seem to be better than what public schools can produce. So this should be possible for motivated parents.
Different kids need different things, and the same kid might need different things each year.
I'm not claiming this is a perfect system. And others have thought through this better than I have.
>Why does a kid even need to attend school if they can go to such a museum with guidance?
Guidance from whom? A human being? Wouldn't that human being thus be a teacher, and the museum a school?
Sort of but not really. The museum is chosen by the parents as part of a curriculum, one designed or children by the parents.
The teacher can be the parent (if they have knowledge in the area) or a homeschool teacher hired by the parent. Parents of homeschooled children with different areas of expertise form co-ops and teach each other's children based on their expertise.
The basic difference is if it isn't going well for the child, the teacher can be fired immediately.
I know this is probably very worrying for Americans because this is going to be your government, but today's headlines about Trump's cabinet picks have me laughing.
New Secretary for Education is, if I go by the headlines, "WWE wrestling mogul" Linda McMahon. I know there was a lot of disdain for his last pick there in the first administration, Betsy de Vos, so I'm awaiting (with a certain amount of glee, I have to say) the reaction to this. It would seem her experience in education was serving on the Connecticut state board of education 2009-2010 and being on the board of trustees for a Catholic school:
https://www.irishexaminer.com/world/arid-41520317.html
"She served on the Connecticut Board of Education for a year starting in 2009 and has spent years on the board of trustees for Sacred Heart University in Connecticut. Seen as a relative unknown in education circles, she has expressed support for charter schools and school choice."
More Catholics in the cabinet? I am delighted to see the Sinister Papist Conspiracy to extend the tentacles of the Vatican into American government is proceeding apace!
https://commons.wikimedia.org/wiki/File:Anti-Catholic_octopus_cartoon.jpg
https://www.reddit.com/media?url=https%3A%2F%2Fpreview.redd.it%2Fus-anti-catholic-cartoons-portraying-the-catholic-church-as-v0-4elnjjf86nnd1.jpg%3Fwidth%3D768%26format%3Dpjpg%26auto%3Dwebp%26s%3D739ca4399e09440f78fff0a25e87a778fb71c67b
Dr. Oz to oversee health insurance? Okay, even I have to blink at that one.
Head of Commerce department could affect us here in Ireland, as (and I have to agree on this one), allegedly he tweeted (or Xed, is that how we should say it now?):
https://www.rte.ie/news/business/2024/1119/1481894-trump-to-nominate-howard-lutnick-to-lead-commerce-dept/
"Lutnick was recently critical of the nature of Ireland's trade relationship with the US.
"It's nonsense that Ireland of all places runs a trade surplus at our expense. We don’t make anything here any more – even great American cars are made in Mexico. When we end this nonsense, America will be a truly great country again. You’ll be shocked," he wrote in a post on X in October."
It is nonsense, and it's not anything tangible (except for the pharma plants here), it's all down to our tax strategy to attract multinationals (like the FAANG set) which have headquarters here and move money around from one bank account to another. The effect on our economy is such that we need to have a separate calculation for GDP to account for the 'phantom GDP' from this.
So I wouldn't be surprised if the Trump administration changes taxes to make it attractive for American multinationals to keep their business at home, and of course this will have a knock-on affect on Irish employment as/if plants are closed or the Silicon Docks crowd shutter their offices. We'll have to see what happens!
I'm curious: what's your view of the worst case scenario here? Like what's the worst possible outcome that a bad DoEd secretary could cause? In my view the DoEd doesn't do anything positive for the cause of education, so it's impossible to do any harm by nominating an incompetent secretary. I think the country would be better off if the entire dept was completely abolished.
The DoEd becomes dysfunctional without having its duties transferred to other agencies, six million college students (mostly from poor or working-class families) lose their Pell Grants, many of them have to drop out of school, their lives are diminished, the nation's supply of skilled labor is diminished, and there's a massive electoral backlash of mostly Trump supporters looking for the anti-Trumpiest possible candidates in the next election.
If you were under the impression that the Department of Education was just a bunch of wokescolds who send out letters to people trying to run good schools and saying "you need to be more woke or else!", then no, that's not how any of this works.
> six million college students (mostly from poor or working-class families) lose their Pell Grants
And? They're still college students. They are affiliated with the left as far as the populace is concerned.
If the economy does go to shit, they have more than enough scapegoats to blame it on. Hell, if left wing people end up leaving in droves, the administration won't even have to rig the next election in order to maintain power!
And why is that a bad outcome? In my view far fewer people should be going to college, *particularly* those from poor and working-class families. Elitism is good in elite institutions. Eliminating Pell grants would go some way to eliminating cost disease and bloat in higher education. Academia has become a lumbering zombie and a propaganda indoctrination zone. Anything that cuts it down to size is good in my view. Education needs far less regulation.
Wouldn't you support an anti-Trump backlash?
< In my view far fewer people should be going to college, *particularly* those from poor and working-class families. Elitism is good in elite institutions.
Well, Wanda, I come from a family with only one college-educated parent, and our finances were very tight indeed when I went to college. My parents were especially honest people, and did not fudge the truth to make us look poorer than we were when they filled out the financial aid forms. Our income and savings were low enough that I got almost a full scholarship for the entire 4 years. I graduated from an Ivy, scored 99th percentile on the GREs and got PhD. Jeez, maybe if you'd had Betsy DeVos's job I'd never have gone to college, and would instead be your hairdresser. Or, like, the lady who gives you Brazilian waxes.
As for other teens whose Pells you'd like to yank: I just looked at some data about four year colleges and their average SAT score.
https://www.univstats.com/corestats/admission/colleges-by-average-sat-score/
For about 35%, the average SAT score is 1200 or above. 1200 places someone at 64th percentile in the group that took the SAT, & that group's average intelligence and academic performance is surely above the average for college students as a whole. I'd say it's someone scoring that well or better on the SAT is not in the in your born-to-be-your-hairdresser (or hair-plucker) group, and can certainly benefit from college-level courses. If you nix the Pell grants of all the kids in those schools, that's a lot of people unable to pay for an undergrad degree they could probably have made good use of. And only 8% of colleges have have SAT scores 1000 (the average score) or below.
Ok, you're the exception. Getting 99th percentile on your GRE's means you have a >120 IQ. You're exactly the kind of poor person that's supposed to be discovered via standardized testing.
>For about 35%, the average SAT score is 1200 or above.
Ok. An SAT of 1200 is an IQ of about 120 which is about the top 10% of the population. That's who should be going to college, not the ~50% that we currently send. I would MUCH rather sacrifice the 35% who should be there for the 65% who shouldn't. And pulling Pell Grants wouldn't prevent them from going, it would just make it more expensive. I want poor, potentially-stupid people to think very carefully about whether their mongrel genes can actually benefit from an expensive education before they just blithely go. College used to be a sorting mechanism. It's not anymore and that loss imposes very reals costs on society. It's incredibly useful for everyone to know where to get smart people.
>And only 8% of colleges have have SAT scores 1000 (the average score) or below.
Ok. That number should be 0. If someone gets a 1000 on the SAT then they shouldn't come anywhere near a college. College is for the top 10-25% of society. Everyone else should just learn a trade. Hairdresser is an excellent example.
<And pulling Pell Grants wouldn't prevent them from going, it would just make it more expensive.
"Let them eat cake."
Actually, the way the colleges set financial aid is by using the data the parents send about their savings and income, and the data parents must give is *very* detailed. There really is no loophole that allows people to look less able to pay for college than they are by having a secret rental property or whatnot. Of course they can lie, but colleges are allowed to do some checking around so that's pretty daunting. Using the data the parents sent, colleges calculate how much parents should be able to pay, and their idea of what people of a given income can cough up is generally more than people themselves feel able to cough up (though most can by penny-pinching. To cover the amount parents will not be expected to pay, colleges give a financial aid package, which is a mix of scholarship (i.e. college eats the cost), loans the parents or kid must take out, and money the kid must earn with an on-campus job. Financial aid packages always set the amount the parents and kids borrow as the maximum in government student loans they are allowed to borrow . Colleges give scholarships to cover the rest, and for kids who qualify for Pells the scholarship includes the Pell grants. So if the Pells are yanked, either the colleges would have to be willing to eat more of the cost or the kid and parents would have to borrow more. However, the family is generally already stuck with borrowing the maximum they are allowed to as federal student loans. The alternative for borrowing more is bank loans. People who are poor enough to qualify for Pell grants, and have already committed to take out $10,000 in student loans for the coming year generally do not qualify for bank loans.
<I want poor, potentially-stupid people to think very carefully about whether their mongrel genes can actually benefit from an expensive education before they just blithely go.
Jeez, Wanda, what even are mongrel genes?
< I would MUCH rather sacrifice the 35% who should be there for the 65% who shouldn't.
You know, I think I agree that people scoring at or below average on the SAT generally can't benefit from college courses. (Although there are college majors in things like hospitality, and when I was on vacation in Mexico I met a young couple who had actually majored in that and were looking like they were going to make a go of it with an eco-resort.). But I find you extraordinarily mean-spirited, and would even if I were not in the 35% you'd rather have doing your hair. It's pleasant to remind myself, and you, that nobody but you gives a shit who you are willing to sacrifice.
I'd prefer the backlash be more "reversion to the mean" and less "pendulum swings back like a wrecking ball", and I'd prefer not to have millions of innocent young men and women thrown abruptly out of school to that end.
And, yeah, among the hip rationalists and contrarians, academia is all cost disease and bloat and indoctrination and signaling. To the *vast* majority of Americans, the sort of social mobility represented by someone from a working-class family going to college and hopefully making a better life for themselves, is seen as a *good* thing. That's the story that made J.D. Vance a household name, but I suppose as an Ivy-league alum we're supposed to write him off as an irredeemably woke leftie or something.
The now-predominately-Republican working class likes that sort of social mobility, because those are *their* kids with a shot at what they see as a better life. The Democrats like that sort of thing because they think they'll get more Democrats out of it. I like that sort of thing because I'm a libertarian who believes in meritocracy. I'm not sure what you are that you don't see this as a good thing, but whatever it is, there aren't enough of you to matter.
And the bit where there's at least a large minority that's annoyed by all the kids going to Oberlin to study Uselessology or Columbia to join Team Hamas, fine, but *these aren't those kids*. Those are mostly the ones from upper-middle-class families who assume they'll be getting an upper-middle-class lifestyle just for having a college degree, and feel guilty about it. The ones who need a Pell Grant to make it through school, know they need to make that education count for something in the real world. So, yeah, pat yourself on the back for having reduced the number of college students, but you're getting rid of the students and your pissing off their parents and they're going to make you pay for that.
A+ comment, no notes.
I think the problem is that university is associated in people's minds with being above average on the social ladder, so the popular sentiment is "if we all can get university education, then we will all be above the average".
Unfortunately, this is not how averages work. If everyone gets university education, then you will need PhD to get a hairdresser license.
So perhaps instead of extending the sweet deal of "get more credentials in return for a lifetime of debt" to an ever increasing part of population, we should go in the opposite direction and reduce the need of credentials. To become a hairdresser, you should need one month of training, at most. Make it so that if you decide that being an unemployed bum or a software developer sucks and you would prefer to be a plumber instead, you will only need two or three months, and you are legally free to start your own business.
Separately from that, we should improve education, and there are many things that could be improved there. But we should not make it legally required in order to get a good job.
What does "if we all can get a university education..." have to do with the policies of the second Trump administration, or with this discussion? Nobody here is proposing to give everybody a university education. There is no plausible outcome where everybody gets a university education in the next four years.
*Some* people will get university educations. Some people *should* get university educations. Not so much the people who are planning to become hairdressers, but definitely the ones who are planning to become engineers, doctors, lawyers, or businessmen, Ms (or Dr?) Tinasky may be of the opinion that those should all come from the middle to upper middle class and the proles should stick to their prole jobs like hairdressing. I'm strongly of the opinion that a good number of our future engineers, doctors, lawyers, and businessmen should come from the ranks of the working class and even some of them from the poor. And I'm pretty sure the wast majority of American voters agree with me on that.
Right now, the way that happens is with loans and grants that come through the Department of Education. I can imagine worlds where it doesn't work that way. Worlds where wealthy philanthropists or large charities or state governments handle that, and worlds where college is cheap enough that an ambitious young man or woman can pay for it with a part-time job. But we don't live in one of those worlds, and we're not going to build one of those worlds in the next four years.
So, if you break the Department of Education, then A: millions of students who I *think* you and I agree really *should* be going to college, will have to drop out, and B: a large majority of Americans will be really pissed at whoever made that happen.
If you've got ideas for substantive educational reform, great, but they're going to take more than four years. And to make them work as well as you hope, you're probably going to want to have the DoEd in good working order to either carry out the reforms or to seamlessly transfer operations to whoever is going to carry out the reforms.
>To the *vast* majority of Americans, the sort of social mobility represented by someone from a working-class family going to college and hopefully making a better life for themselves, is seen as a *good* thing.
Those people are wrong and I don't think that the federal government should be in the business of laundering their delusions. The "everyone should go to college" idea is just a hedonic-treadmill effect. The government shouldn't fund that.
If you're a working class kid who has class-bucking IQ then fine, suck it up and take on some debt. College isn't the express-lane to upper middle class that it used to be, so low-class families should legitimately stop and think about whether it's the right choice. Eliminating federal grants is something that would encourage a more rational cost-benefit analysis among the consumers of higher education. College was designed to be for the top 20% or so of the population. Anything more than that is just a waste of resources.
<If you're a working class kid who has class-bucking IQ then fine, suck it up and take on some debt.
You seem to be very underinformed about paying for college. Did you think Pell grants cover everything? Nope, they average about $5000 per year. And the average kid with a Pell grant ends up with a 15% *larger* debt, approx $26,000 total, than kids who did not qualify for a Pells, presumably because the Pell does not fully make up the difference between their funding and other kids'. So be of good cheer, Wanda, the Pell kids are doing plenty of sucking it up.
I can't tell what Peter is trying to say, but: people do want social mobility, of the type that comes from being capable of doing things that pay well enough to fund their entry into higher social strata. What they *don't* want is the type of social mobility that only comes about by bending the knee to the people currently in the top stratum because they have arbitrarily seized that stratum and have changed the entry criteria away from being capable of producing actually valuable things. Possibly taking Pell grants hostage in the process.
> And why is that a bad outcome? In my view far fewer people should be going to college, *particularly* those from poor and working-class families.
Definitely because nobody wants social mobility. The real problem is the rich bit dim, and what is needed is a reduction in certain courses but that’s less likely to happen.
Seems like Trump is in an awkward place. He put a number of normal people with regular credentials in his cabinet last time, and not only did he get lots of flack about it anyway, but lots of the people he picked tried to foil him. So he's highly incentivized to bypass all the normal picks.
I don't know what people expected differently, to be honest.
"Huh, everyone who is reputed to know what they're doing thinks I'm a moron and tries to derail my plans. But since I'm obviously infallible, I guess I should go with people who everyone says are clueless next time". Peak Dunning-Kruger right there.
Maybe, or maybe there really is a "deep state" of people with similar views across party lines and they hate Trump for his opposition to their one-party rule.
I'm not saying that's necessarily true, but I also don't think we can rule that out. I'm troubled by how many things Trump got criticized about that Obama did before him and Biden did after. Kids in cages being an obvious and well known case, but also Biden continuing Trump's tariffs and things like drone strikes against enemy leaders.
Let me rewrite the first sentence from a different perspective: Maybe there really is a group of public servants with a deep knowledge of how the machine runs and commitment to keep the state running regardless of party affiliations, who really hate Trump for his blatant stupidity and disregard for reality and morality.
Of course, that's the optimistic spin, and the reality may be somewhere in the middle. But I really believe that any complex organization needs a lot of people with specialized knowledge, and replacing them will lead to dysfunction. The question is, "is the current state of things so rotten that it's worth tearing big parts of the system down and having effectively nothing in their place for a while?" Apparently, in the opinion of Trump, Musk and others, it's worth it. My guess is that they're going to replace somewhat corrupt but functioning systems with ones that are both dysfunctional and utterly corrupt... but we'll see.
That is a really good question, and I also do not have the answer to it.
I do have a problem with leadership in a democracy ignoring the well-intentioned pleas of the population. This includes people on the left (Bernie supporters) as well as people on the right.
Many of the problems are not exactly subtle or hidden. That outsourcing and deindustrialization has gutted the entire rust belt is not news to anyone. That the leadership class seems to think of this as somewhere between fine and good is a serious problem. What's an economy for? They seem to think that number-go-up is enough. Serving the needs of the population is a better answer. If number-go-up also helps the population (as it did in the 50s) then everyone can be satisfied. If instead it helps people in certain fields to a massive extent but most people are hurt or gain little, that's going to cause problems.
I don't think Trump is a cause, but a symptom. I've been listening to working class Republicans hating on their elected Republicans for well over 20 years. It was only a matter of time before someone showed up to do what those people actually wanted, instead of more wars and more international trade. If the solution ends up being bad, which is entirely possible, then that still doesn't fix the underlying issues. The only thing that will fix the issues is those so-called experts to take the needs of the lower classes seriously.
I assume that those working class Republicans you talk about hate on their elected Republicans in between elections, but vote Republican at election time.
The Biden Administration passed the CHIPS act and the Inflation Reduction Act, and we saw the election results. It’s possible that the policy would have been a political winner with better messaging; I personally don’t have any expertise in political messaging so it’s hard for me to say. It’s also possible that the policy helped politically, but not enough to overcome other factors which allowed Trump to win. But it suggests that addressing deindustrialization is not a politically winning policy.
Indeed, you go on to write that these Republicans voted against more wars and more international trade. The first has nothing to do with economics, and the second is different from the issue of outsourcing and deindustrialization gutted the rust belt. If you put tariffs on the import of raw materials, that makes American manufacturers less competitive, so it reduces international trade both by reducing the import of raw materials and by reducing the export of manufactured goods. Great if your goal is to reduce international trade, bad if you care about American manufacturing.
Based on polling done in 2019, Bob Altemeyer wrote, “The biggest reason by far that people supported Trump was their level of prejudice, but some (not many) relatively unprejudiced subjects approved of him because they thought they were prospering thanks to him.”
https://cdn2.mhpbooks.com/2020/08/Authoritarian-Nightmare_Appendices.pdf
I see what you mean, but I am not sure the wishes of the electorate can be fulfilled in a straightforward way without unintended consequences hitting them hard years or decades later. My understanding is that the bleeding-out of parts of America's industry was an inevitable side effect of globalization (i.e., free trade and Pax Americana), which not only lifted hundreds of millions of people worldwide out of poverty, but also brought about relative peace, a ridiculous pace of technological progress and, last but not least, the possibility of bying cheap stuff from China. Rolling it back may or may not lead to workers in the US having plenty to do (depending on how it's handled), but it very likely leads to an overall lower standard of living in the US and probably, within a few years, lots of wars leading to a thorough disruption of global trade, neocolonialism, biblical-level famines and general unpleasantness which eventually impact everyone, including the workers in the US. (Peter Zeihan paints the full picture in "The end of the world is just the beginning"). People who have worked for decades in the State Department probably understand this; the guy whose economical platform consists of "I LOVE tariffs" maybe doesn't, and neither do many of the voters who can't find China on a map.
Curious which picks of his you think got a lot of flack while being normal. At least some of his picks were quite popular - James Mattis for SecDef was confirmed 98-1 and H.R. McMaster for National Security Adviser was confirmed 86-10, for example.
EDIT: Other popular choices:
- Shulkin for VA, 100-0
- Nikki Haley for UN Ambassador, 96-4
- John Kelly for DHS, 88-11
- Elaine Chao for Transportation, 93-6
- Perdue for Agriculture, 87-11
- Ross for Commerce, 72-27
He got burned in two different ways. Careerists who sailed through confirmation tended to stymie his plans. Anyone else got pilloried as unfit even before they got in the job.
Would we really expect Biden, Harris, Obama, or Bush to keep playing by the informal rules when those were the results?
I mean, if your choice is between "qualified people who will easily be confirmed but won't do whatever I tell them to" and "widely-criticized people who will do anything I want", I think MOST people would wonder whether the first group actually are doing a better job than the second group. There's more to cabinet roles than rubber-stamping the President's agenda, and surrounding yourself with a bunch of yes-men could be pretty self-defeating.
Sure, but Trump ran on a platform of doing things differently than other politicians. His options are to either follow the consensus like every other president and thereby betray his voters, or put people in his administration that aren't part of the consensus.
I think there's a lot of room for us to discuss which examples are "non-consensus" and which are "unqualified cranks".
E.g. Rex Tillerson for State was definitely not a direction I would have gone in, and I would argue was pretty meh for the department, but at least was an international businessman and probably falls into the "not willing to do whatever Trump wants" category. RFK, Jr. is a lunatic conspiracy theorist with the opposite of qualifications who could do astonishing damage to public health.
Yes, I think this is a factor in Trump's more controversial picks for his next cabinet. It's pretty clear that he felt burned by some of the "normal people with regular credentials" in his first administration, so this time he's putting a premium on reliable allies and friends. Trump has been close to the McMahons for decades, so no surprise he's turning to Linda again, this time in a larger role.
Conservatives have for years wanted to dispense with the federal Department of Education; it's not immediately clear whether Linda McMahon is being put in charge to try to carry that out. Certainly the Project2025 guys want it to happen as described in their big blueprint document.
The McMahons, husband and wife, are actually a helluva shared life story which some competent writer could make a compelling book out of some day. A couple since high school, early in their married life they were on food stamps and a bit later went bankrupt; neither had been born into any money. Despite which they became very successful and will soon celebrate their 60th (!) wedding anniversary. Frankly they are in some ways what Trump cosplays as....Linda held a lesser appointment during the first Trump administration, and once won a party nomination for a Senate seat. So she's not now a complete rookie to the political/governmental big leagues.
Lutnick is someone who's pursued an admirable adult path after a rough start (his mother died of cancer during his last year of high school and then his father was killed by medical malpractice during his first week of college). His qualifications for Secretary of Commerce are not obvious but then again that job has become largely toothless anyway; his appointment isn't much different than our longstanding shitty practice of naming large campaign donors to be our ambassadors.
Meanwhile though I am not at all laughing at Trump naming his literal personal defense attorneys to the #2, #3 and #4 positions of our Justice Department. _That_ is a new one which not even the likes of Tricky Dick or Slick Willy or Harding's handlers ever tried.
McMahon converted from Protestantism to Catholicism. Vance is also a Catholic convert. You'd think Protestants would be bothered by losing so many of their future elites. Or maybe they're proud of being a religion for low-class people.
Mass immigration from third world means internal white difference are no longer relevant. No one cares if you are an Irish (American) or Swedish (American) or just a good old Anglo. When a third your university class consists of recent arrivals from China and India, another third children of Latin American immigrants, you are just "white". The massive secularization starting in 2000s means the same process has happened with religion.
> no longer relevant
...Yet. They'll run out of non-white scapegoats eventually.
I think they're just happy with this country returning to true Christian rule, even if they're not the sect in charge.
> So I wouldn't be surprised if the Trump administration changes taxes to make it attractive for American multinationals to keep their business at home
I thought the point of Ireland was that it was the cheapest (tax-wise) European state to be in? If I recall correctly, there is a significant tax advantage (or requirement?) for big tech companies to keep a European bank account if they’re doing business in the EU…
That's what Trump's guy is complaining about, and others before him. The EU told Apple to pay us €14 billion in back taxes, and our government had to be *forced* to take the money, they tried putting it off because that's our one advantage, our low tax regime for foreign investment.
https://www.breakingnews.ie/business/apple-tax-billions-begin-to-roll-into-state-coffers-1691491.html
As I get older it seems that cats, in general, like me more. When I was a teenager, even friends' cats would give me a wide berth. Then progressively catkind got more neutral about my presence. Now it's a semi-regular occurence that cats actually come up to me and ask to be petted. Even completely strange cats, in the street. I don't think my behavior toward them has changed (I'm not aggressive but reserved. Since I'm mildly allergic, I avoid touching them). Is it because they see fidgeting as scary and I got more immobile with age? Has anyone else noticed it happens to them too?
It seems to be cats specifically and not pets in general.
A primate grin is NOT welcoming to a cat, quite the reverse as one is showing teeth. Most cats are also averse to wide open looking eyes, especially staring at them, as they view these as a challenge. Also, outstretched human arms tend to alarm them, especially when these move abruptly.
A cat smile is a slow blink and gently narrowing eyes. So maybe your eyes have become apparently less open and alert looking as you age, possibly behind glasses you never wore as a teenager, when cats were avoiding you. Either that, or you regularly eat toast or muffins for breakfast. Cats can overlook most worrying quirks of human anatomy if the humans smell slightly of butter! :-)
You may have cracked it, I have a drooping eyelid and now that you mention it, in all three of the last instances the cat approached me from that side. Never saw (heh) the pattern.
They can sense that you're infected with Toxoplasma gondii.
Damn, and apparently it's incurable.
Why isn't it a bigger deal here that the benefits of scaling AI larger seems to have hit the ceiling? Shouldn't that be the biggest AI story in the world? Why aren't chip and utilities stocks crashing?
I know they say blah blah blah we got other ways to improve AI --- but fucking shit --- the way to improve it so far has been scaling on bigger data sets. Why should believe another way of scaling is going to improve things all of a sudden?
Well the "AI Explained" channel had a maybe-it-will-maybe-it-won't segment :-)
It was already priced in.
Nine months ago people around here were going "Why isn't the market reacting to the fact that AI will soon be godlike and/or destroy us all?" But people outside this bubble never really expected that to happen.
Because the NVIDIA etc stock bubble hasn't burst yet.
Stipulating for the moment that AI has hit something ceiling-ish: Hype sells. There's stories to be told about hype, and people want to tell, hear, and be part of those stories. Blood also sells, and that includes the green-tinted financial sort of blood. The interregnum where the hype has worn thin, but the last round of Bigger Fools haven't given up and sold out yet, that's just kind of meh and boring.
And around here, we've basically got people who believe AI is wonderful and awesome, and we've got people who believe AI is an awesome and terrible danger, and we've got people who think the first two groups have been talking about AI way too much. If it turns out that AI stagnates at "meh", then for both the wonderful-and-awesome and the awesome-and-terrible groups, talking about AI would mean talking about how they were wrong. Nobody wants to do that. And the third group has had all the AI-talk they care for so they're not contributing more.
Sooner or later, either AI will break through the "ceiling", or the bubble will burst, and either one of those will be a big deal here. In the meantime, all is quiet on the AI front. Enjoy it while it lasts.
>Sooner or later, either AI will break through the "ceiling", or the bubble will burst
A third, "meh"er possibility is that LLMs stagnate - but the current level winds up being good enough for a lot of applications (which may include enshitification of customer service). Remember how long it took for businesses to work out just how to use personal computers intelligently!
edit: (from the https://www.theintrinsicperspective.com/p/ai-progress-has-plateaued-at-gpt article):
>if AI paused tomorrow, people would be figuring out applications for decades.
I do think we can rule out a near-term stagnation of machine learning on _trustworthy_ data, e.g. the protein fold prediction. That doesn't have the hallucination problem that predict-the-next-token has.
Yes, this has all the vibes of 1999 internet bubble. Both things will likely come to pass: the bubble will burst, and a useful machine-learning infrastructure will be available to used for many more years.
Yes, that sounds plausible. And, when the internet bubble burst, the stocks came down, but the internet didn't go away, as if the whole technology was a bad idea, so I agree with your expectation that machine learning applications will persist.
Hmm... There _have_ been bubbles where the underlying technology really was a bad idea. E.g. https://www.orau.org/health-physics-museum/collection/radioactive-quack-cures/emanators/national-radium-emanator.html the radium craze around 1920-1930ish... https://en.wikipedia.org/wiki/Radium_fad
I'm pretty sure LLMs stagnating, even at a useful level, still results in a spectacularly newsworthy bursting of a bubble that's currently valued on the expectations of LLMs very much not stagnating.
That sounds plausible. Many Thanks!
edit:
<mildSnark>
2030: "And one of the more famous points of interest in Silicon Valley is the Nvidia stock crash crater, formed in 2026..."
</mildStart>
I'm kind of surprised that the people most concerned with AI singularity haven't attacked it more from an energy conservation perspective. The amount of energy required to keep improving AI is truly immense, and it shouldn't be hard to make a basic environmentalist argument against using that much energy on AI, even putting aside concerns with possible AI superintelligence. I'm surprised that more people in *this* social circle haven't done so, at least from what I've read so far.
Probably most people here are already concerned with anthropogenic climate change. And much of that climate change comes down to energy consumption. I have some very leftist friends that are highly anti-crypto currency precisely because of how energy-intensive crypto mining is (at least from what I've heard). Well, it sounds like AI is becoming even more energy-demanding than crypto-mining is.
If you're concerned about anthropogenic climate change, and if you're concerned about AI superintelligences, then it seems there's now a great way to tie them together and attack both at the same time. It seems odd to me that I haven't read this argument put forward more.
As I understand it, OpenAI's GPT was developed essentially by simply ploughing through every recorded utterance since the dawn of time through 2018 or whenever it was. On that assumption, If energy is a limiting factor then presumably further progress can be made only by structuring the data better, into a network of more self-contained units, and enhancing these independently.
If we draw a rough parallel with programming, the first operating systems were presumably masses of messy machine-specific assembler, but over the decades these have become far more modularized at the same time sophisticated.
People absolutely do complain about the amount of energy AI uses, including here.
Well, when the same people are pushing for electric cars, it looks transparently dishonest to be demonizing electricity consumption in non-Woke applications.
Pushing for electric cars happens when the choice is between many little engines burning fossil fuels to accomplish some task and using centrally generated electricity to do it.
When instead the choice is between doing a task and not doing it at all, yes, the same people will push for the latter option.
No-one's arguing for clockwork+petrol AI; Nvidia is not, last I heard, building giant Babbage engines for running LLMs, though I do admit that would be kind of awesome; and no-one's putting petrol engines in their data centres.
Meanwhile, the people pushing for electric cars do also very much tend to push to reduce car use overall.
All seems consistent to me. Where's the dishonesty?
>Where's the dishonesty?
In pretending that the electric cars won't require electricity generation and transmission infrastructure enhancements along the same lines as for AI or any other non-Woke-favored use.
>Meanwhile, the people pushing for electric cars do also very much tend to push to reduce car use overall.
That _is_ a fair position, and characterization of their position. I generally oppose it, but it is at least honest.
> I generally oppose it, but it is at least honest.
I do feel there is a massive disconnect between people who live in very densely populated areas where owning and using a car is misery and people living in suburban sprawl and/or rural areas, where a dense public transport network would not see enough use to be feasible and so there is no sane alternative to car ownership; people in each of the two groups who have not lived in the other kind of place find it very hard to get in each other's headspace.
But that is by the by.
Many Thanks!
>so there is no sane alternative to car ownership
Yup, I'm in one of those areas.
> In pretending that the electric cars won't require electricity generation
Does anyone pretend that? I've only ever heard it as a MAGA talking point, trotted out while ignoring whatever the interlocutor actually says; much as you just did to me.
Many Thanks! I've usually seen it as the _absence_ of discussion of all the electricity generation and transmission requirements in some electric car advocacy piece. Look, I've seen Greens block _transmission lines_ , the national equivalent of an extension cord, let alone power plants. Now, these may not be the exact same people who are pushing electric cars, but, as an overall political group, in aggregate, they look hypocritical.
Frankly, for the training phase part of AI development, solar or wind could potentially be used. Training is an incremental process. Nothing intrinsically horrible happens if the process runs during daylight hours. There is an economic trade-off due to having the capital cost of the chips lying fallow at night - but if we _really_ are in or near a regime where energy costs dominate, this should be ok. Solar _without_ storage is one of the cheapest, possible _the_ cheapest, source of electricity.
If we _aren't_ in a regime where energy costs dominate, then these energy discussions look like a dishonest cudgel being used to bash AI which wokesters actually oppose for _other_ reasons.
I don't think most people in AI -- myself included -- think we have hit a ceiling. Everyone in the world is compute capped. We need more chips, and more energy. But chips and energy are continuously getting better and cheaper. It shouldn't be lost on anyone that tech basically lobbied the government to make nuclear a green energy category and starting buying a bunch of micro reactor time as a result. That should give you a sense of what the industry thinks of as our current limitations.
Related: there are several good papers on scaling laws that show that, following a power law, exponential improvements in compute and data together result in linear improvements to models. Anecdotally, however, linear improvements in models lead to exponential improvements in acquired abilities and outcomes. This is part of why everyone is compute capped. Every single GPU is already going towards training the largest models possible, they need more to scale more.
(Inb4: Data is not the limitation -- Google has tons of data they are purposely not using, no one has the compute capacity to really use images and definitely not video or audio.)
I realize he isn't in AI, but why is Erik Hoel wrong in this post:
https://www.theintrinsicperspective.com/p/ai-progress-has-plateaued-at-gpt
Some excerpts:
"Employees who tested the new model, code-named Orion, reportedly found that even though its performance exceeds OpenAI’s existing models, there was less improvement than they’d seen in the jump from GPT-3 to GPT-4.
If this were just a few hedged anonymous reports about “less improvement,” I honestly wouldn’t give it too much credence. But traditional funders and boosters like Marc Andreessen are also saying the models are reaching a “ceiling,” and now one of the great proponents of the scaling hypothesis (the idea that AI capabilities scale with how big they are and the amount of data they’re fed) is agreeing. Ilya Sutskever was always the quiet scientific brains behind OpenAI, not Sam Altman, so what he recently told Reuters should be given significant weight:
Ilya Sutskever, co-founder of AI labs Safe Superintelligence (SSI) and OpenAI, told Reuters recently that results from scaling up pre-training—the phase of training an AI model that uses a vast amount of unlabeled data to understand language patterns and structures—have plateaued."
What to make then, of the claim that researchers have unlocked a new axis of scaling by giving the models time to “think”? For this is what the entire industry will now rely on, both in practice but also to keep the sky-high hype alive.
I think this new axis of scaling will matter far less than people expect and will not keep the train of improvements chugging along anywhere near the previous rate.
First, this new type of scaling appears limited to certain domains: it makes the models no better on English language tasks or provides only minimal improvements for questions about biology, for instance. Second, there are good reasons to believe that this sort of scaling will be much more expensive: linear improvements for extreme sums of money."
As tech journalist Garrison Lovely wrote about it on Substack:
If you have a way of outcompeting human experts on STEM tasks, but it costs $1B to run on a days worth of tasks, you can't get to a capabilities explosion, which is the main thing that makes the idea of artificial general intelligence (AGI) so compelling to many people…. the y-axis is not on a log scale, while the x-axis is, meaning that cost increases exponentially for linear returns to performance (i.e. you get diminishing marginal returns to ‘thinking’ longer on a task). This reminds me of quantum computers or fusion reactors—we can build them, but the economics are far from working.
But I think people focusing on price or the domain-specificity of improvements are missing the even bigger picture about this new supposed scaling law. For what I’m noticing is that the field of AI research appears to be reverting to what the mostly-stuck AI of the 70s, 80s, and 90s relied on: search.
Re Hoel's essay:
Saturation along a straightforward axis is indeed always bad news.
I'm not as pessimistic about the combination of LLMs + search as Hoel is.
LLMs have always seemed to me to be like system 1 thinking - pattern recognition on steroids. And search is similar to system 2 thinking, articulating possibilities, then evaluating their relative merits.
I'm _not_ a worker in this field, so my gut reaction is ill-informed, but these _look_ complementary, and might still be a pathway to AGI.
I do think we can reasonably say that _some_ sort of artificial neural net must be able to reach AGI, since _we_ are neural nets, and _we_ exhibit general intelligence. But, this doesn't say that the _current_ artificial neural net interconnection _architecture_ can reach AGI, and exploring multiple alternative candidate architectures, at many millions of dollars of training costs per experiment, might be a deal-breaker :-(
I'm not one to go against Sutskever, his track record is certainly better than most. That said, I'd also be curious what, exactly, he's doing over at SSI, and how much he stands by how Erik is interpreting his words.
More broadly, 10x-ing the amount of compute and data that is fed into a GPT is no small task and I'd be somewhat surprised if anyone has gotten there. Is Orion 10x bigger than GPT4? The latter is almost 2 trillion parameters trained on 25000 GPUs using 13 trillion tokens. Is Orion 20 trillion parameters trained on 250000 GPUs using 130 trillion tokens? I'm really doubtful.
But such a model may well be linearly better than GPT4, in the same way GPT4 was linearly better than GPT3. There's just massive engineering challenges that need to be addressed first. I don't think Sutskever can solve those engineering problems with SSI, the company is not big enough. Note too that Sutskever was sorta probably pushed out of OAI, he didn't leave out of some scientific disagreement. Meanwhile, Microsoft (OAI) or Google might be able to actually hit the level of scale necessary for a real GPT5, and they are both certainly trying (again see the nuclear power expenditure). But getting your hands on 250000 GPUs (or some number of TPUs for Google) is tough! So imo it remains to be seen that the scaling hypothesis is truly dead. And my own circles of AI researcher friends sure seem to think there's more juice to squeeze.
>And my own circles of AI researcher friends sure seem to think there's more juice to squeeze.
I wish them luck! There may be other parameters to push, but it would certainly be bad news if (reasonably) straightforward scaling has plateaued.
There will always be people willing to throw money at fancy new tech. And the big companies have set out their stalls with AI versions of their own, from Microsoft to Google, all producing consumer versions to "try this and let it help you".
So there is too much money already spent on this to stop, or for share prices to fall. The faith is that AI progress will keep going, or even if it doesn't, that widespread adoption of it will keep the profits rolling. Internet of Things Mark II.
It's not that big of a deal, because even if the scaling laws do hold, they imply that you need exponentially increasing amounts of power and compute to eke out marginal gains in ability. What good is an AGI if you need one data center and one power plant to replace one human employee? Or even 1,000 humans?
I think it's pretty obvious that a pure transformer approach can't lead to true AGI, because we have yet to find a way to train models on small data sets, and there's only so much you can put in the context window. For a lot of tasks, there's just not enough data available to train an LLM on, you have to "learn on the job", and a pure transformer can't do this.
OpenAI's o1 approach is much more promising and shows that scale alone isn't everything. I'm somewhat sure that LLMs will be part of the first AGI, however.
"we have yet to find a way to train models on small data sets"
Humans have GI (General Intelligence), and I think THEY don't actually work on small data sets. Humans take in an awful lot of data points every second. Just the eyes have about 120 million rods and cones, each of which are collecting data many times per second.
Scaling up a process is linear, no matter how you do it. Economies of scale can reduce per-unit costs, but linearly. LLMs are an important milestone, being able to do things computers weren't expected to be able to do, but it will take another kind of technological leap to have geometric growth in AI.
We are born wired to pay attention to language, and to use the language we learn to make sense of experiences. We are wired to experience it as carrying meaning. We have a special brain area set up to organize and remember the language we hear, and find regularities and remember them. We are born with a craving to make sounds with our mouths, and to imitate sounds we hear. we are born preset to transfer what we know to other modalities —
writing, sign language, foreign languages. All that gives us an enormous advantage over a machine that has to figure out stuff from scratch.
Agreed on everything except your last sentence. Yes, we have, e.g. Broca's area.
But the language processing is the part of AGI that is a _solved_ problem at this point. Probably inefficiently solved, but solved. As you showed with your experiment with the pendulum problem, they are still getting very simple physics (what bumps into what) wrong. And I keep seeing them get simple chemistry questions wrong. But these aren't language problems.
Yeah, you're right about my last sentence. In fact it's odd to be denigrating AI about language, when it seems like that's the thing where it comes closest to being our equal. When I talk with GPT4 now, I don't even think about simplifying, just write my questions or comments exactly as I would here, and it understands fine. What I had in mind, but did not say, is that our inborn ability to master language is set up so that learning language and learning about the world link up easily, in fact they are sort of two sides of the same coin. Here are a couple small cute examples. : When my daughter was learning to talk we had 2 affectionate cats who hung out with her a lot. One day she was touching something soft -- I can't remember what -- and she said "it's kitty." And she didn't mean it was a kitty, she meant it was soft, as cats are. Learning the textures of things, and learning words for them, were part of the same process. Not knowing the word *soft* she felt free to invent a word by extending the meaning of a word she knew, and knew I knew. Here's another example, sort of a cousin of the first, I think from when she was 3: We were in a Starbux and slow jazz was playing, and I asked her if she liked the music. "It's whiny," she said.
So the hard-to-duplicate thing about our species isn't that we are set up to learn language, and so can do it with way smaller samples. It's that we're set up to learn langauge *and integrate that* with learning about world. So the way we know what pendulums do comes both from things we've read or been told and from handling objects.
Many Thanks! Yes, the LLMs have a lot of trouble with things that we learned with our senses, and that we, as you said, _linked_ to language.
>It's that we're set up to learn langauge and _integrate that_ with learning about world.
Yup. And a good chunk of how humans learn is by acting on the world, both physically and socially, and seeing the reaction. As you said:
>When my daughter was learning to talk we had 2 affectionate cats who hung out with her a lot. One day she was _touching_ something soft -- I can't remember what -- and she said "it's kitty." And she didn't mean it was a kitty, she meant it was soft, as cats are.
[emphasis added, and also agree about the generalization she did]
Reading enormous amounts of text can substitute for _some_ of this, but LLMs clearly have problems in these areas, and it isn't clear how far into something like robots-who-can-play we would need to go to get around them.
"We are born wired to pay attention to language"
I dispute this assumption. We are born with awesome pattern-matching brains, which recognize sounds, gestures, differences in pitch, and myriad other things, and determine ways to use some of these things to communicate with others. Language is learned behavior.
Machines, on the other hand, are nothing but Chinese rooms so far. Even the idea that they interact with language must be programmed into them. Any dataset can be interpreted any number of ways, and it is up to the programmers to choose which ways.
It’s well known that our brain is set up to attend to and learn language. Really. If you doubt me, google something like “is human brain optimized for language.”. iIf you still doubt it, read the links.
I don't think you understand me. Yes, the brains perceives the components of language. But "language" itself is for communication, which is something the brain determines for itself. We do know an awful lot about how the brain handles such things (https://en.wikipedia.org/wiki/Language_processing_in_the_brain).
Consider a thought experiment of taking some people as babies, and raising them without teaching them any known languages. They will likely develop their own languages to interact with each other, and making mouth sounds will probably be only one part of that. If a baby were abandoned in an isolated area and somehow survived, it would not develop any language capability, until it conceived the desire to communicate, perhaps with animals it befriends.
So I'm saying that the brain is CAPABLE of language, and is good at it, but it isn't really programmed in any way for language specifically. A bicycle is designed for travel. It could be placed into a device that converts the energy of motion into electricity. Even though it generates electricity well, it remains designed to convert a person's energy into wheel rotation.
> Humans take in an awful lot of data points every second. Just the eyes have about 120 million rods and cones, each of which are collecting data many times per second.
I don't think this argument works, because almost none of that sensor data is used for text-based learning. When a human reads the definition of a previously unknown word plus one example (corresponding to maybe 50 tokens), they're able to use it from now on. It doesn't matter that the eyes are collecting Megabytes-equivalent of data while reading that short text, because that sensor data still only contains about 50 tokens worth of training data.
An LLM is able to temporarily "learn" to use a new word this way, as long as its definition is still in the context window. However, to permanently learn this word (i.e., to modify the model itself), a lot more examples and training data are required. Maybe a sufficient amount of synthetic training data can be generated by the model itself from the definition in the context window; however, I don't consider this a "pure" transformer architecture anymore, but rather a system with the LLM as one of its components.
You're thinking of learning like a computer or data scientist. Living things don't think that way, and though it may be inefficient, it is certainly effective. LLMs analyze text because they have a format of data analysis programmed in that allows symbols to be correlated in that way. Brains, however, figure out HOW TO CORRELATE THE DATA, and which data to correlate.
In other words, which pieces of data should be included in analysis? We make the decision for LLMs. Imagine giving LLMs long strings of bits to analyze, say 2^100 bits long. What is it to make of it? We could have encoded a bunch of pictures in a certain format, or encyclopedias of informational text, or some format of blueprints of various buildings, or recordings of various sounds. A brain figures out what the sensory information means, but a LLM must somehow be "told".
>and which data to correlate
That's essentially what the "attention" layers in the transformer architecture do. And the adjustable parameters in them get optimized during training, along with the other parameters, so, yes, LLMs do learn
>which pieces of data should be included in analysis
> Why isn't it a bigger deal here that the benefits of scaling AI larger seems to have hit the ceiling?
I expect its still denial
maybe also delaying the consequences
of this view becoming the norm.
What does Europe do? It used to be a fun place to vacation, but now they get angry at us when we go there. What purpose does Europe serve anymore?
Come up with solutions that are then ignored by other countries.
Well, it's a pretty good place to store all the Europeans so they don't come over here and be angry at us.
Also, there are still plenty of places in Europe where people don't generally get angry at Americans. Thought I'm certain they'd be willing to make an exception for certain Americans.
Did you think the purpose of Europe was American tourism. The vast majority of tourism in Europe is from Europe.
A few cities are feeling over congested but that is it. Mostly tourism is a very important part of the European economy
As other respondents said, I think the basic premise is mistaken (or very context-dependent at best). I can think of a few unserious answers as to the purpose of Europe.
1) Give Americans places for cheaper (yet chic) vacations where you can drink tap water.
2) Showcase how to properly be left-wing (think of how many meanings this sentence can have!).
3) Give the World Spirit some psychological drama for this grand story of stories that They’re realizing.
4) Give more data and customers to tech companies (there are more Facebook users in Europe than in the US).
5) School of life for the US “greatest generations”. Why don’t Americans cook up another big European war again? (As everyone knows*, the UK only got the atomic bomb because France did. You can still rekindle the 900-old rivalry!)
*Well, that’s what actors playing top-level UK bureaucrats in a comedic 1980’s TV show say in-character anyway. Why would they lie?
Wonderful list!
What makes you think they get angry with us when we go there? I've been there multiple times in the last 12 months, I didn't encounter anyone who seemed angry at my presence.
I think they mean there's a bunch of people in social media saying "argh tourism is bad", and anecdotally I've seen graffiti in more than one place I traveled to saying "refugees welcome, tourists go away" (granted in the rest of Europe, not in England).
I think this sentiment is just people blaming bad economic conditions on a scapegoat, and it's probably a minority of people. If tourism actually decreased many European cities would actually suffer.
I have read that high profile destinations like Venice are getting overwhelmed by large scale group tourism from countries like China whose citizens aren’t well socialized for tourism (yet). Also, social media drives large numbers of young people to take their selfies in the same Instagrammable locations, causing bottlenecks.
Yeah, people used to complain about American tourists but that was before Chinese tourists were a thing.
There's people on social media saying literally anything you can think of at any given moment. I'd save the worrying about people being mean to you when you vacation in Europe for when they are, actually, mean to you when you vacation in Europe.
Most of the anti-tourist attitudes are probably aimed mostly at other Euros.
Chances of random Euros being "angry at Americans when they go there" have never been particularly high as far as I know, but is probably lower now than during the Iraq War.
The purpose of countries isn't to serve Americans. People live there!
As an American, I can only say ?!?!?!?!
From one source (https://www.reddit.com/r/Jokes/comments/8rmsx2/the_un_decided_to_do_a_worldwide_survey/):
The UN decided to do a worldwide survey, with just a single question. The only question asked was: "Would you please give your honest opinion about solutions to the food shortage in the rest of the world?" The survey was a huge flop.
In Africa they didn't know what "food" meant. In Eastern Europe, they didn't know what "honest" meant. In Western Europe, they didn't know what "shortage" meant. In China, they didn't know what "opinion" meant. In the Middle East, they didn't know what "solution" meant. In South America, they didn't know what "please" meant. And in the USA, they didn't know what "the rest of the world" meant.
I love this joke!
This made my day. Thank you for sharing the joke!
Who is “us”? I recently went to Europe and they weren’t angry with me.
Why does Turbo Tax have a near monopoly on tax software? Is it mostly due to:
1. Marketing and branding
2. Quality, ease of use
3. They got in early with the IRS and have a sweetheart deal
4. Other
The software actually works. I have all sorts of negative feelings about Turbo Tax in specific and regulatory capture in general, but they found me a deduction once that will cover all the money I'll ever pay in Turbo Tax fees for the rest of my life.
It might be partly mindshare, and I say that because I'm a TaxAct (H&R Block) user, but yes TT is the one everybody knows.
It's not like Turbo Tax is comparable to Microsoft Office applications. Excel and Word have inertia because they are the standard we share with each other in the business world.
We don't need to share our Turbo Tax with each other, right? So why do people keep turning to it instead of a cheaper alternative?
They can prepopulate much of my data as a returning customer, and I can typically do my taxes each year in less than an hour. For me, it's not worth the search cost to find something cheaper to save $100.
Also, I've heard it alledged that TT lobbies for more complex tax laws to help maintain their confuseopoly position in the market, but not sure of this.
Regardless of your politics, what are people doing to prepare for the incoming administration? Buying crypto or real estate to get around inflation? Getting all your vaccinations done ahead of schedule in case access is disrupted? Buying organic food in case food safety is disrupted? Nothing in particular?
I bought oil and gas stock index and partially sold long-term (20+ years) US government bonds index. Also I've started to incrementally increase my S and P 500 allocation.
Practicing my "sympathetic yet ironic detachment" face in case anyone IRL tries to talk to me about it.
all of those things sound like massive overreactions to me, but if they help you sleep at night then you do you.
My default assumption is that Trump's second term will be a lot like his first (hopefully without the pandemic at the end).
> "all of those things sound like massive overreactions to me"
The "every-day-it's-worse" cabinet picks really seem to indicate otherwise, but hopefully you are right
Epstein for Secretary of Education! Oh, he's dead? OK, that old girlfriend of his who's in jail, how about her? And I know it's a bit weird, but how about a dung beetle for Sec. of Health. And what about that guy who wore a Viking helmet in the Senate for Sec of State. Fearless, you know?
Dung Beetle? Quietly takes care of shit - perfect resume for any cabinet position I say.
Buying Trump buttplugs for Christmas presents
https://www.etsy.com/il-en/listing/1358286625/donald-trump-butt-plug-desk-office?click_key=23bd5be96665820e34c760f08cd20328434190ee%3A1358286625&click_sum=31266222&ref=sold_out-5
I'm being trolled. But that's OK.
So you WANT Trump up your ass? Interesting.
Up yours.
Ive was watching multi stage whole house scale water filter prices
I'm considering buying several 2.5 gallon containers of glyphosate and triclopyr.
* Bought Series-I bonds up to the purchase limit for myself and my wife.
* Decided to take the saver's credit this year instead of realizing capital gains, as the Trump tax cuts aren't going anywhere for at least another four years.
I'm buying lots of handguns and knives
To use them on who, yourself or others? Because if it is the latter, I highly suggest against it! You'll be completely fine if you don't do anything stupid. And you're not an illegal immigrant. Or a legal immigrant who's black. Anyways, even if that was the case, fighting back will just get you killed faster. Best to just stay put and watch the fireworks.
I'm preparing for a life of crime for when the economy goes to shit.
Ahahaha. Well, I hope you realize that the police and military usually outlast the economy. Their loyalty is the number 1 priority for any government, after all.
Any trustworthy analysis on why the final Selzer poll in Iowa (which showed Harris *ahead* by 3 percentage points) wound up so wrong (Trump won by 13 percentage points or so)? Selzer has a long and impressive track record, and it doesn't look like there was a dumb spreadsheet entry error or anything of that sort. Did the poll just get massively unlucky, on a level analogous to "we got p<.005, but the null hypothesis was in fact correct; this was the 5-in-a-thousand chance"?
One of the wackier conspiracy posts about the election I've seen is that Selzer's poll was correct *but* Musk and Trump hacked the voting machines and stole the votes, which is why the results of the poll didn't tally with the election results.
That's some religious level of faith in "Ann Selzer is always right" there.
On the other side of the wackiness gap, I've also seen suggestions (but how serious versus tongue-in-cheek I have no way of assessing) that Selzer was going to retire anyway, so she decided to burn all her credibility to get the result she wanted. Release a poll showing Harris with a massive lead, hopefully discourage all the grubby MAGA voters from turning up and energise the Brat Summer set to turn out and be part of the victory.
I don't believe either set of alternatives, but I do lean towards "Selzer was getting the results she might have subconsciously wanted" as one reason why the polling was so out of sync. Even pollsters are only human and have their biases.
The analyses I've read emphasize that her polling method was unusually "hands-off"; that is, she adjusts for very few factors. I found an old 538 article saying merely "age, sex, and congressional district". It turned out that, in this election, there was an effect that strongly confounded that uncontrolled analysis - I would suspect education as the culprit, given that the college/no-college split widened this year, and that education correlates with polling response rates. There was probably also some nontrivial amount of random error, though I suspect it didn't have to be on the order of thousands-to-one.
I think if there's an aberrant thing here, it's people being overimpressed by her track record - overcorrecting based on a small set of dramatic events, starting in 2008. No shade to her, really - she's always been transparent about her method, why she likes it, and its limitations - but it should have been clear that her method wasn't magic, and it was always vulnerable to this. While it avoided pitfalls others fell into in other cycles, as soon as there was a really significant split which wasn't substantially correlated with her sole controls of age, sex, and district, her methods were going to yield results divorced from reality. It's a shame it happened so dramatically when she was planning to retire anyway, ending her career on a sour note.
> I think if there's an aberrant thing here, it's people being overimpressed by her track record - overcorrecting based on a small set of dramatic events, starting in 2008
I don't know if they were overimpressed by her track record, so much as they were overimpressed by the fact that she was telling them what they wanted to hear. I swear that 99% of the people telling me how reliable Selzer's polls were on November 2 had never heard of the name Selzer on October 31. She went from obscure specialist to household name and back again faster than anyone since... well, I dunno, I'm sure there's other examples but I've forgotten them all.
There were outliers on the other side too which didn't get nearly as much attention. I remember a few polls showing Trump winning New Hampshire, New Jersey, New York.
Prior to this election, at least Nate Silver was impressed with her, and his judgment counts for something in my book. (A lot of the commentary around the latest poll specifically may have been Democrats hyping news they wanted to here, but readers of Nate Silver would have been familiar with Selzer for a while.)
Yeah I expect Nate Silver to be familiar with Ann Selzer, I was just bugged by all the random civilians who were suddenly telling me that Ann Selzer was the bestest pollster ever despite the fact that they hadn't heard of her a week earlier.
Someone, I think maybe Silver or a commenter on his site, mentioned that the crosstabs were crazy. Way too many women, especially within certain subfields.
This is helpful - thank you!
This. The only thing I'd add is that she's been famous in large part because she was historically willing to buck the trend, and while she got lucky the last few times (and is actually a good pollster), being willing to buck the trend does mean you'll also be loudly incredibly wrong sometimes. So her being so visibly off here is in some ways correlated with her being impressively right in the past.
This is the man Trump wants in charge of Medicare and Medicaid:
https://x.com/ShadowofEzra/status/1858995616161890670
I was expecting to see people criticize Dr. Oz as a pick, I was not expecting to see someone criticizing him on the grounds that he was supportive of a trans kid on his show.
Watch the entire clip. The "girl" barely remembers being a boy. Then the mother (it's ALWAYS the mother) "reminds" her that she didn't like getting her hair cut and she compliantly gives the "correct" answer that it was because she was a "girl not a boy." Munchausen syndrome by proxy.
I did watch the clip, which is why I described it the way I did rather than the inflammatory way the tweet does. The mother says they didn't know anything about being transgender until the pediatrician diagnosed it, and I certainly wouldn't accuse someone of making up the story for attention based on a one-minute clip of them.
And regardless of what you think of the mother, Dr. Oz just seems generically polite and validating here, in the way I'd expect most TV personalities to be. Were you expecting him to yell at the girl and tell her she's a boy?
"Were you expecting him to yell at the girl and tell her she's a boy?"
Ideological Turing Test failed.
Rhetorical questions are attempts at ITTs now, apparently.
Beat those strawmen!
For anyone who doesn't want to follow a Twitter link, FYI it's Dr. Oz and that's not a joke.
The year is 2030 and the US is in a civil war. Given those facts, what do you think the most likely steps were that led to the war?
Trump's pick for Minister for Technology makes a statement officially recognising Vim as superiour to Emacs.
There isn't any plausible scenario for a US civil war in the next 6 years, but I'll try:
Trump goes full slash-and-burn against the Federal government: FBI is disbanded; HUD, Dept of Education eliminated, others radically downsized; Federal funding is pulled from all Universities that have any DEI or grievance-studies departments. This creates massive economic dislocation as significant streams of federal money dry up. At the same time AGI is created at OpenAI and they declare explicit anti-administration political intentions as the state of California aligns with them. The administration begins a serious crackdown on illegal immigration and sends National Guard troops to detain them. Coastal cities declare themselves immigration sanctuaries and refuse to admit federal troops. Trump freaks out about OpenAI and considers it a national security priority to have the newly-minted AGI firmly in control of the federal government. OpenAI refuses and California aligns with them. There is open rioting in the streets as political protests grow unruly so Trump sends Federal troops to California to a) restore order b) detain illegals and c) secure OpenAI. Democrats win the midterms on the back of a 2008-level recession and rampant inflation. They use their control of Congress to impeach Trump on completely baseless politically-motivated charges. The vote is close but he's removed from office amid rumors of unscrupulous practices by the Democrats (which are actually true - they threaten the families of several key swing votes with violence in a very "wink wink" organized crime sort of way) so Trump refuses to leave office and appeals the impeachment to SCOTUS which rules in his favor. Progressives declare SCOTUS a GOP-apologist kangaroo court and refuse to accept the verdict. Coastal progressive states collectively declare Trump an outlaw and refuse to cooperate with the Federal government. Military leaders are divided along regional lines and bases in progressive states formally refuse to acknowledge Trump's legitimacy. Hilarity ensues.
Disgruntled fanatics on the right side of history suddenly discovered the Second Amendment was in fact right and dropped all their calls for gun control so they could tool up and riot in the streets.
Being leftists, they were of course fissiparous, which is why you now have the People's Independent Democratic Republic militia duking it out with the Socialist Democratic People's Front terrorist cells and both of them making and breaking alliances with the Democratic Republic of All Folx Peaceful Protest (And Suitcase Nukes) Units, themselves considered by analysts to be in danger of breaking up between the All Folx, the Y'All Folx, and the "Damn it, it's spelled 'folks' Grammar Communist Cadre" factions.
The rest of the country just grills, drinks beer, and goes to sportsball games 😁
I can see it.
Had or in?
If its had, trump and vanced, possibly rfk, got hit by an american cruise missile; and even your most feminists pro abortation wine aunt far leftist just yields to the right wing and theres just purges, 95% of the miltrey defects or stands down and there's a constitutional crisis; the right wing is in chaos without a figure head and violence being ovisous action but mostly unplanned. I think there are just elements of the leftwing that drink that much koolaid and are just that dumb. If the left wing starts the war in a big obvious way be short and brutal, the neonazis could probably get segregation, dont be an accelerationist.
In, it probably is a slow build up, both sides maintaining plausible deniability; the american empire collapsed, there been a tax protest for a few years and the finance system of the world is just a mess, the world is rapidly decentralizing, mass suicides, and half the effort is shitposting prosuicide messaging to the other side on a fractured no longer global internet; which is very effective especially after the international spying databases get opened up for blackmail; hard line political segregation is happening, with whatever the current rate of people fleeing California is 10x'ed and *every* state. Abortion clinics get bombed in red states, churches in blue, both in purple.
AI leads to superpowered scissors statements and/or superpersuader effects. As a result, you can get any of these:
* Almost every person hates almost every other person.
* Polarization gets taken to the extreme.
* People randomly get sucked into online rabbit holes and emerge as people with drastically different views and values than they had before.
* AI which can convince people of anything just by talking to them, so the first movers get massive cults behind them, and use them however they see fit. Conversation is acknowledged as dangerous, so communications are shut down, exacerbating previous disputes.
>AI leads to superpowered scissors statements and/or superpersuader effects.
Hmm... Kind-of an AI-driven zombie apocalypse?
Interesting.
Big protest of something or other, police show up to break it up, shots are exchanged, and a celebrity of some kind is killed in the crossfire, prompting both sides to go all in. Who's the American Francis Ferdinand? I'm going with... Weird Al Yankovic. The assassination of Weird Al Yankovic leads to civil war.
When you can't find a damned foolish thing _in_ the Balkans, find a damned foolish thing _from_ the Balkans (I like Weird Al, FWIW)
I like it.
We find out the entire calendar system is significantly wrong, and the date previously thought of as 1863 is in fact 2030.
You guys are no fun sometimes.
Define civil war?
If your definition involves neat lines of blue or grey coated men carrying muskets, then I would think the most probable steps involve invention of an ASI which then placed us in a Matrix style simulation, deciding for its own ineffable reasons that the *nineteenth* century was the pinnacle of our civilization.
This is the inadvertent correct answer.
The US will get into a civil war over the proper definition of "civil war".
The Navajo Reservation decides to secede from the United States.
The US Government doesn't want to acknowledge it, but also doesn't want to step in and violently put it down, so there's an awkward state of technical civil war happening.
Would it mean anything legally for the Navajo Nation to secede from the US? Does anyone know?
I'm doubting that "technical civil war" means anything. I think civil war can only exist if a reasonable person thinks it does or some standard like that.
It's a civil war fought primarily with technicals.
https://en.wikipedia.org/wiki/Technical_(vehicle)
(The best kind of civil war!)
Well if there were a legal means for the Navajo to secede then it wouldn't be a civil war any more. It didn't "mean anything" legally for the South to secede from the US, or the US to secede from the British Empire, but that didn't stop it from happening.
You can imagine various low-level skirmishes going on that turn it into a more real war if you like, to raise it above the threshold where you could say "hmm I guess this does really count as a civil war" but not above the threshold where the US military simply comes in and stomps everyone.
I don't think this is likely, but a tiny civil war that just counts as a civil war seems like a much more likely scenario than a massive war of all against all.
Hmm... https://en.wikipedia.org/wiki/Utah_War - Does this count as a tiny civil war?
I'm going to say "Trump refuses to leave office and gets a few generals to agree with him on that", or "After Trump leaves office the Democrats charge him with treason and execute him, leading to a popular revolt", or "Congress passes an Australia style firearms ban." But honestly, I can't see any of those as being likely to happen.
Thanks.
Of course civil war isn't likely. But given that it's not likely and given that it happened anyway, I'm curious what sort of scenarios people think are most likely.
I’ll go with a breakaway state firing on Fort Sumpter.
Has anyone done a Scott Alexander style “more than you ever wanted to know” about seed oils?
My prior is that the whole thing pattern matches to every nutrition fad: some interesting but not terribly conclusive studies challenge conventional wisdom (in this case, that linoleic-acid is good for heart health) and trigger a current year bogeyman (inflammation!), then pop culture health influencers amp it to 11 and insist they’ve discovered the one magic bullet that explains all our health woes. Bonus points if you can blame it on modernism and double bonus if you can spin a conspiracy theories about BIG FOOD and their effort to POISON us for PROFIT. Until a few years later none of it really pans out, nutritionist update slightly to “maybe eat a bit less of that if you’re sensitive” and the health influencers move on to the next demon food.
But I’d be happy to approach on objective summary with an open mind.
Heather Heying and Bret Weinstein have a sort of intro-level explanation of why seed oils are bad, or rather, why one ought to infer at a layman level that they're probably bad (they presumably have empirical research that supports the layperson reasoning): seeds, unlike other parts of a plant, do not want* to be consumed. They want to grow into new plants. To help with that, they make themselves non-nutritious to animals, relative to, say, fruit, which *does* want to be eaten in order to carry the seeds elsewhere to spread the plant's lineage. In general, we can expect a part of a plant to be nutritious if (a) eating it would help the plant's lineage and (b) the plant wasn't artificially engineered into "forgetting" what helps it.
They have a book called _The Hunter Gatherer's Guide to the 21st Century_, that I think goes into some detail on this.
*"want" here is in the evolutionary sense (Heying and Weinstein are evolutionary biologists) - shorthand for "evolve in the direction of".
> seeds, unlike other parts of a plant, do not want* to be consumed
The counter example that disproves this is that some seeds (e.g. coffee beans) do get eaten and pooped so they can spread.
"Consumed" here means in the sense of being dismantled and used for nutrition. Simply passing through a GI tract doesn't count toward this. (Indeed, an animal won't want to eat things simply to pass undigested through it, unless those things come with some other benefit. The plant, by contrast, would be perfectly fine with this, and encourage it for its seeds.)
Evolution is too wacky to consistently apply anything like this. There are all sorts of healthy and edible plant parts that aren't the fruiting body; carrots, onions, potatoes. In fact, the fruit of the potato plant is poisonous to humans while the tuber is not, in direct contradiction.
I agree, but will also broaden it in that people aren't the only beings to eat fruits and vegetables. For example, deer eat and enjoy poison ivy, and many birds love eating seeds of various kinds. And anything produced by a plant, if not sprouting into a new plant, will be recycled by other living beings, such as bacteria and fungi.
>while the tuber is not
Well, there is _some_ solanine in potatoes. Green potatoes can have enough to be toxic https://medlineplus.gov/ency/article/002875.htm
Well, I agree evolution is pretty chaotic. To take your examples: all three of those vegetables are very aggressively bred from the original wild stock, so to some extent, all bets are off. Or alternately, we say these lineages have "adapted" to being spread by the humans who are functionally symbiotes, and indeed, the human-bred varieties grow in vast numbers.
The wild carrot, aka Queen Anne's lace, birds nest, et al., is a wildflower with a woody root. The seeds are apparently edible and aromatic, but maybe only if cooked, and I can't tell if they're nourishing.
The onion is virtually only existent as an artificial cultivar; the wild varieties (there are several) have bulbs that animals will eat, but the bulbs are not the seeds.
Potatoes grow from their eyes, which are nodes in the tuber. Aha! ...But an animal would have to dig for that tuber. Certainly, they do. I'm honestly not 100% sure what's going on there; maybe the eyes aren't digestible, so gophers et al. end up just spreading them the same way tomatoes get spread.
Hmm. I've eaten onions and other edible stems like leeks growing in the woods, which afaik are wild cultivars. But when eating a fruit, whatever eats it also ingests seeds which are excreted later, propagating the plant. That doesn't work for tubers or stems, since eating them prevents the plant from sprouting. Potato tubers sprout from eyes, which contain the same inedible toxins as the potato fruit. But this can also be avoided by just eating the potato before the eyes sprout.
I'm skeptical of seed oils, but...wouldn't that same logic apply to animal fats? No animal *wants* to get eaten, after all.
An animal doesn't want to die, but doesn't much care what happens to its fatty tissues after it does. Either it has surviving kids or it doesn't, and evolution is mostly going to work on "don't get killed until you've had plenty of kids".
A plant, evolutionarily speaking, *does* care what happens to its seeds after it dies. Or even while it still lives but the seeds have fallen/been harvested. So there's reason for evolution to work on "please don't eat my seeds".
Whether this effect is strong enough to have rendered seed oils subtly unhealthy for human consumption. is questionable. Evolutionary biology is quite good at conjuring just-so stories that are probably directionally true, but not at a significant level. And I'd expect that if evolution *had* significantly optimized for "please don't eat my seeds", the result would be something dramatically different than "really tasty oil that probably gives you heart disease in twenty years".
Good point! I hadn't considered that.
I think the idea there is that animal fat is serving a role for the animal that grows it. Ideally, an animal wants to be as tasteless to predators as possible; however, it also wants to survive times where food is hard to find. Fat is the easiest way to do that, and there are other ways to avoid predators, so animals grow fat, but also claws, quills, sensitive ears, legs that move them fast, etc.
Properly, we ought to apply the same idea to plants. I suspect a seed can only be so unnourishing before it can't even make a new plant, so that's the floor, and the plant has to explore other methods of surviving to the next round as well, and sure enough, while they haven't tried running away, they do grow spines, poisonous leaves, bark, and other tricks.
So then my question would be whether any of these tricks are enough to make a plant just not "care" how good its seeds are to eat, but that ends up functioning on whether it costs energy for a plant to defend its seeds that could be instead channeled into something else that helps it thrive. I imagine a plant wants seeds with enough stored energy to grow that new plant, but not so much energy that animals seek it out, so no matter how I slice it, it makes sense to me that a plant would send some of that energy into animal-repelling tech, but maybe not as much into parts of itself that aren't quite as critical - or even sending that into something that *attracts* animals that will help it spread, such as birds and bats and insects.
What’s the actual mechanism? That sounds like a just-so story.
Animals also do not want to be eaten, and yet their meat is tasty and nutritious.
I don't know their account much beyond what I described above, but if I tried to lay it out based on what I know of evolution:
Plant seeds will be nutritious or not. If they aren't, then animals will adapt to not eat them, and the seeds will survive, and so will the lineage. If they are nutritious, animals will adapt to eating more of them until they go extinct, or adapt to be non-nutritious or repel the animal with something that leaves that animal worse off, encourages the animal to not digest the seeds, etc.
Ergo, plant seeds "want" to be less nutritious. Ergo, any plant that's evolved on this planet sufficiently long (and hasn't been extensively fiddled with by geneticists or breeders) will possess seeds that aren't nutritious - or are as non-nutritious as the plant can manage while still permitting germination, or will be coupled with toxins.
It isn't one sided though - any animal that wants to eat seeds will be more successful if it can digest seed oils. And humans in particular seem to be pretty good at eating a wide range of plants that have evolved to avoid getting eaten (tolerant of caffeine, capsaicin, etc.)
I think the idea here is to distinguish which things humans have adapted biologically to eating, from which things they've adapted to technologically (either humans to them, or them to humans). The evolutionary biologist's argument is that the former process enables us to be nourished and not poisoned by what we're choosing to eat through good old fashioned natural selection; the latter goes too fast for that to work, and we could easily damage ourselves faster than the biological adaptation process can do its work to acclimate.
One possible answer to this is to speed up acclimation with technology, too. I haven't seen Heying and Weinstein address this possibility specifically, but based on my experience following their Q&A sessions, I'd guess their answer would be that this is possible in theory, but currently, technology doesn't know how to do this. Moreover, they might maintain that the human body, like any lifeform, is so complex that it's likely to outpace any technological solution in the foreseeable future - including the parts of our bodies that get harmed by technologically engineered food. (This is consistent with a rule of thumb I *have* heard them say, which is, when in doubt, the fewer the processing steps performed to make your food edible in the short term, the better it probably is for you in the long term.)
Hmm. Fruit is the only part of plants that is “designed” to be eaten (plus or minus nectar to attract pollinators), and most herbivores do not exclusively eat fruit. Ergo I don’t think the logic fully holds.
At any rate, seeds must contain some degree of nutrition, else they would not perform their function of fueling the newly germinating plant.
Lots of seeds are not easily digestible (to survive being eaten with the fruit) but that doesn’t mean they are necessarily inedible, nonnutritious, or toxic.
The seeds I think of as indigestible are the ones with a hard coating that looks designed to survive going through a GI tract or just being avoided by a rummaging animal. So, yeah.
I agree seeds have to be at least *somewhat* nutritious - while you responded here, I was saying as much to other replies. The idea there is that there are multiple ways for that plant to defend its seeds. One is to reduce the reward to the animal by reducing the nutrition. Another is to raise the cost with seed coatings, and toxins.
In this case I agree with Nassim Taleb, considering that we have a lot of oils we've been using for millenia, why become a guinea pig testing new stuff? it's almost impossible to prove that modern seed oils have no long term adverse effects
The problem is that it's almost impossible to prove that [insert words here] have no long term adverse effects. If this is our standard nothing new can be ever done.
Not at all. If something has been consumed for millenia we can observe long term effects. Also one could argue that genuinely harmful substances would not pass the evolutionary filter but I'm less sure about this argument.
“If something has been consumed for millennia we can observe long term effects.”
I don’t think this is at all obvious. How exactly would we? How would this experiment be designed?
Basically agreed. If something has very visible adverse effects, and it is very obviously linked to consumption of a given food (e.g. group A and B are identical except group B consumes one extra food, and regularly keels over ten years before group A) then maybe...
Generally, though:
- We only started to look at disease patterns systematically with John Snow's work on cholera in 1854. Earlier than that - well, people _were_ dropping dead quickly from cholera, it _was_ from water (from a particular source) that they consumed, and no one isolated the cause before then.
- Even for things we have been consuming for millennia, the dose may change. Sucrose has been consumed for a long time, but the _amounts_ that e.g. the USA population consumes are recent
- It is damned hard to isolate the effects of _one_ food (short of having people drop dead immediately after consuming it). Given different groups which have all sorts of differences, including many differences in diet, attributing illnesses (particularly long term illnesses) to particular causes is excruciatingly difficult
Even more damning, the Romans used lead for plates and plumbing for - at least - centuries, clearly unable to detect its toxicity. Like you said, only obviously toxic foods where the link between consumption and illness is immediate, are easily identified.
The State of California warns of the long term adverse effects of [living in the State of California].
Literally! See, e.g., Prop 65.
As being tautologically the sort of person who reads Scott Alexander, I was hoping for something more than the naturalistic fallacy.
I can't add much to what Michael Watts wrote below, that's pretty much what I think
On a related note, while answering this I learned thanks to wikipedia that "naturalistic fallacy" is also a philosophical term that has nothing to do with appeal to nature as an argument
Perhaps I was being imprecise to call it the “naturalistic fallacy”, but whether it’s that, an appeal to nature, an appeal to the precautionary principle, an isolated demand for rigor, or whatever, I think the essence is the same - you’re just stating a prior to prefer more “natural” stuff absent (impossible) proof of the absence of harm from newer less natural stuff. And I was hoping for a more rigorous evaluation of the “seed oils are bad” claim.
Besides, you’re a guinea pig either way because you don’t eat or live like we have “for millennia”. This whole thing started when scientists thought they had (and maybe still have) pretty good evidence that a reliance on saturated animal fats plus the modern lifestyle does have adverse effects - the “for millennia” safety doesn’t apply because we mostly ate a lot less and got killed by other stuff before heart disease from too much tallow would have been a big problem.
Yes, I'm a guinea pig in many ways, but it doesn't mean that I want to experiment with seed oils when a perfectly good time-tested alternative in the form of olive oil exists in my neck of the woods
You're right that it's a prior. But if you want to argue that it's wrong, perhaps you should show that there are numerous traditionally used foods that turned out to be harmful (because we know of many modern ones that were later recognised as harmful). The only one that came to my mind was alcohol, Claude added betel nuts and agave syrup but when I pressed him on it he had to admit that agave syrup is also a recent invention. So it looks like the only undeniably unhealthy traditional foods are those that have psychoactive effects.
>But if you want to argue that it's wrong, perhaps you should show that there are numerous traditionally used foods that turned out to be harmful (because we know of many modern ones that were later recognised as harmful). The only one that came to my mind was alcohol, Claude added betel nuts and agave syrup but when I pressed him on it he had to admit that agave syrup is also a recent invention.
https://www.samitivejhospitals.com/article/detail/colon-cancer-risk
tldr: Broiling producing PAHs, mold toxins (the article mentions in garlic) nitrite preserved foods with N-nitrosamines
What's the relationship between "considering that we have a lot of oils we've been using for millen[n]ia, why become a guinea pig testing new stuff?" and the naturalistic fallacy?
The assumption that we should be skeptical of new stuff because it’s new (and by implication, leas natural) and that we are “guinea pigs” if we consume it is pretty textbook naturalism.
That's in your mind. The argument is perfectly sound as stated - things we've been doing for thousands of years are unlikely to hurt us. There's no assumption about what is or isn't natural.
Oh, well if you declare it to be correct, how can I object?
Life has a 100% fatality rate. Objectively lots of stuff we do and have been doing a long time is harmful. Objectively lots of stuff we haven’t been doing for thousands of years is helpful to our health and longevity.
At the heart of it, despite the process for getting there, there is nothing really novel about the constituents of seed oil - just the relative composition of the various fatty acid types. Linoleic acid, the supposed bogeyman, is literally essential to human life, we’re just eating *more* of it than we have historically. In a more general sense we *have* been eating heavily processed seeds (cereal grains) for millennia, literally since the birth of agriculture.
So I guess a steel man version of “this is a major change in our historical lipid balance, proceed with caution” I could buy. But that’s a *prior*, not evidence (and the actual evidence seems to show a poor global correlation between seed oil consumption and obesity rates - obesity is much more strongly correlated with GDP).
To the extent Scott has something of a soft prior for Chesterton’s fence barring other evidence, and a general agreement that something about “processed food” is bad, fine (although he seems to share my frustration that nobody can really pin down scientifically what that is). But his article on saturated fat (linked in this thread) is ultimately pretty dismissive of key aspects of the seed oil theory.
> and trigger a current year bogeyman (inflammation!)
This is such a weird bogeyman. Inflammation, like fever, is the reaction you have to something going wrong, your mechanism of putting it right again. The expectation for artificially reducing existing inflammation has to be that it will make your situation worse, not better.
Steelmanning, I presume that the argument is that we want to remove the cause of chronic inflammation, not that we want to suppress the inflammation as a symptom.
In the specific case of seed oils, Don P. is correct. The theory is that a component of the oils (specifically linoleic acid) is a direct cause of inflammation (by one of its metabolites I think?). Basically, “stop consuming this thing that your body feels the need to attack”. Nothing to do with inflammation suppression per se.
You seem to be commenting on a much more general trend.
That's definitely untrue; all the energy is going toward suppressing inflammation itself.
I've asked my mother (a doctor) about this question for fevers, and her take is that we suppress fevers because, in the event that the fever was actually necessary, we can fall back on hospitalization to deal with whatever the fever was going to prevent.
For inflammation, I don't think even that much thought has gone into it; nobody knows what the inflammation is responding to.
Followup question: let's suppose I accept that seed oils might be bad and are probably worth eliminating just in case. What oil should I use in my cooking instead? Olive oil has a low smoke point, and things like coconut or avocado oil have a strong taste.
My experience is that avocado oil has an extremely mild taste, certainly milder than olive oil. Speaking as an avid home chef, avocado oil is my go-to daily oil. For certain use cases, I also keep olive oil (basically just for Italian), bacon grease (richness and searing steaks) and peanut oil (high heat for deep frying) on hand.
butter
Butter is low smoke point. You can turn it into ghee and it works great for frying.
I like the idea of going back to animal fats! Suet and lard and dripping, as well as fats rendered when roasting lamb and poultry:
https://greenpasturefarms.co.uk/tallow-dripping-suet-whats-the-difference/
https://www.telegraph.co.uk/food-and-drink/features/why-top-chefs-love-beef-dripping/
Ghee for the vegetarians. The vegans will just have to eat their food cold 😁
Careful! If the pigs ate a lot of seed oils in their feed then there lard will be full of the same kind of fatty acid chains. Ruminants, on the other hand, reprocess everything they eat into new fatty acids. So tallow is safe, lard not necessarily.
Depends on which theory you subscribe to about why seed oils are bad. If you think unsaturated fats are bad, go with lard or ghee. If you think PUFAs are bad, go with safflower oil… a seed oil. If you think a low omega 6 to omega 3 ratio is important, go with canola oil… a seed oil. Uh oh.
As far as I can tell, anything with high omega 3 is going to have a low smoke point.
Lard is very unsaturated compared to butter, ghee, tallow or coconut oil.
Safflower oil is like 70% linoleic acid, a PUFA, and widely regarded to be the most dangerous one. I'm not sure why you think it's low PUFA.
Most seed oil avoiders don't care so much about the omega 3 to 6 ratio, mostly they try to avoid linoleic acid and other unsaturated fats full stop. Butter is what most people cook with.
I’m seeing conflicting versions of fat breakdowns on safflower oil. I was going off a graphic showing it as having a very high percentage of monounsaturated fat.
I think what’s going on is that there are recently (GMO?) high oleic acid varieties of safflower oil being produced.
In any case there does not seem to be a 1-1 correspondence between “seed oil” and “highest linoleic acid” content, at least among plant derived oils, though a couple of the highest linoleic acid varieties are indeed from seeds. Which I thought odd.
Oh, and you’re right about lard relative to beef tallow, but the poster I’m responding to rejected avocado and coconut oil as being strongly flavored, and lard is more neutral tasting IMO.
Avocado oil being strongly flavored? This is strange, it's actually pretty neutral.
Scott himself did a post that was obliquely about that on his old blog. https://slatestarcodex.com/2020/03/10/for-then-against-high-saturated-fat-diets/
It mostly focused on high-saturated fat diets, but as a part of that addresses polyunsaturated fats (which mostly come from seed oils in western diets, as you likely well know) as a potential cause of obesity and illness.
Thanks. I have a vague recollection of actually reading that post when it came out, but since the current obsession is all about “seed” oils, I guess it didn’t trigger my memory.
Have we really exhausted our exploration of ivermectin? One thing at a time please.
If ivermectin will allow me to eat an occasional bag of Doritos in peace, then pass the damn horse dewormer.
https://dynomight.net/seed-oil/
Is the closest, it basically does a high level literature review of studies in seed oils vs not and essentially takes them at their word. It's likely not as good as something from Scott, since I think it's not trying to figure out every possible interpretation from data / figure out if there are other confounders, but I basically trust this like 10-20x more than people making assertions about "the science", which I imagine is what young otherwise be getting.
Thank you, I thought that was very good. Then again I would, since it basically confirms my priors.
Has anyone here used Duolicious? I'd expect it to be "just like every other dating app, except possibly even more of a sausage festival," but correct me if I'm wrong, please.
I tried it. Girls there were a lot more internet culture, which is good if that's your thing. It was also so much of a sausage fest that I hardly ever got replies. Much lower response rate than other dating sites. Chicks mentioned having thousands of messages to get through. This was closer to launch, so maybe the situation has improved since.
I took the SAT four times in my life and remember each score I got. I also remember my GRE score.
If you ask someone what they got on the SAT, ACT, or GRE, and they took it fewer than 10 years ago, but they respond they can't remember their score, how likely is it that they got a bad score and don't want to admit it? 50% likely? 90% likely?
Data point: I got a very high score (1598) and my default in conversation is to claim that I can't remember. Sharing a score that high is interpreted as bragging and can get very awkward.
I don't think I've ever been in a conversation where people talked or asked about SAT scores. I remember mine, but I never tell anyone because I assume no one cares.
I wouldn't take that as evidence either way. I personally barely looked at my scores bc I knew I'd get a good score. Like GlacierCow, I do remember my percentiles though. An obviously better metric, in my opinion and way easier to remember.
I never took any of these American exams, so that must mean I don't have an IQ at all.
See https://www.astralcodexten.com/i/136151229/problem-only-the-smartest-people-report-their-sats , which finds that people who "don't remember" their SAT score have, on average, six points lower IQ than those who do.
Something I've been asking about for a while, from that link:
> The Biggest SAT → IQ Conversion Site Is Wrong
>> This man has directly converted the SAT percentiles to IQ scores, which is not what should be done. Tests like the ACT and SAT correlate with IQ at about 0.8-0.85
>> Instead, the ideal way to do this is to take the percentiles from the current versions of the SAT and then convert those into z-scores and then regress those z-scores by the mean by the estimated regression coefficient.
I don't see how this can be correct. When we see that the SAT correlates with IQ tests at about 0.8, we also note that that is the same correlation we see between different IQ tests. But we don't regress the results of other IQ tests when reporting the tested IQ given by those instruments - if you take a Wechsler test and your reported IQ is 120, then your tested IQ is 120 regardless of whether I'm secretly thinking about Wechsler or Raven's. Why would we regress results only for the SAT?
THANK YOU for giving me a direct, useful answer to my question. Looking at the other responses, you're a rare breed.
How is that a direct answer? It tells you literally nothing about the question you asked. It does tell you that people who don't remember their SAT score are stupider than people who do, but that doesn't speak to the odds of forgetting versus not wishing to report.
> you’re a rare breed
Isn’t it why we come here at all?
The Rightful Caliph is a shining example to us down here grubbing in the muck of the comments section!
I don't even remember if I took the SAT, or if it was some other test. Anyway, this depends on what you're counting as "a bad score".
Every time I remember my SAT score, it creeps up a little higher. What does this mean?
Adjusting for inflation?
It means you took it underwater and it developed the properties of a fish story.
Your memory is getting better. Obviously.
I don't remember my score exactly (been more than 10 years since then) because I don't even know what the scoring standards are anymore (various SAT scoring system changes over the years). But I do remember that I scored in the 97th percentile and I remember this because it's the closest analogue I have for IQ I have. What does this make me?
1) How many years ago did you take it?
2) Do you have absolutely no memory of what your score was, or could you confidently narrow it down to something like "around 1400" or "in the low 1300s"?
If I want to guess it was something like 750 on math, a little less good on the reading/writing. Looking online I was in the 2400-score-years so it gets confusing. I think I could confidently narrow it to something around 2100 (+-100ish?) but I had to look online just now to remember what that meant so I think I've contaminated the example.
I took it in 2012.
Yes but do you remember the fastest time you ever got on a 100m sprint?
People who base their self worth on test scores are likely to remember, people who don't are more likely to forget.
Why would you even ask someone that, if you're not in charge of college admissions or something? Just so that you can feel superior if their score is lower than yours? I don't blame anyone for weaseling out of that sort of game.
I don't ask people what their standardized test scores are in normal conversation.
I guess one of the first things Ted Cruise asks someone he meets for the first time is, “What is your IQ?, if they don’t know he asks “What were your SAT scores?”
He’s not a very well liked man in the Senate.
It depends on how many bits of information are involved? I also remember, even though it’s been a lot more than 10 years, but it’s not hard since I only have to remember one bit of information.
But also, why do you care? I’ve never had anyone ask me my sat score after high school, and I would definitely judge negatively someone who asked me that today.
I took the PSAT and the SAT, once each, and remember both my scores but not which one goes with which test. They're pretty close, but less than perfect, so it doesn't much matter.
Technically SAT scores are ten times higher than PSAT scores.
You can compress the data by only storing `total number of points lost across all standardized tests.'
That's not an approach I would recommend when the scoring scales are so different in absolute terms. It's similar to suggesting that one mile plus one meter is two. You'd need to convert both scores to the same scale first.
I also wouldn't recommend it as stated even for multiple instances of the same test; your compressed score will get worse and worse the more you take the test.
Works just fine as long as the compressed number is zero
So SAT = 10 * PSAT? So P = 0.1?
Are you sure that's reportable?
One thing I don't understand is why it's constitutional for Trump to run for office in 2024. Amendment 22 says you can only have two terms, and Trump won in both 2016 and 2020.
There's no evidence he won in 2020 that didn't immediately fail when analyzed with some scrutiny. If this was a joke it was mildly funny though lol
I guess on a serious note, Trump running for office in 2024 is an implicit admission of the loss in 2020. That is, if we assume Trump cares about the Constitution. Well, never mind!
It's an implicit admission he lost in 2020 iff he does not also plan to run in 2028.
Are you Attorney Paul Gordon, submitter of the hilarious legal troll challenge making that argument to the Maine Secretary of State? That objection was considered and rejected: "Application of the term limit turns on whether an individual has _actually been elected President twice_, not on beliefs or assertions about that fact."
https://www.maine.gov/sos/news/2023/Decision%20in%20Challenge%20to%20Trump%20Presidential%20Primary%20Petitions.pdf
He won on November 3, but not on January 6, and that's what matters.
Interesting, the plain language of "No person shall be elected to the office of the President more than twice" suggests it's the election that matters, not the ratification.
The election by the president happens by the electors, with the election of *those* electors being a separate matter, insofar as I've understood. The electors conducted the election of the president on Jan 6.
When precisely do you believe the election is COMPLETE, and the person is considered to be elected? When the last polls close on election night? When the ballots are all completely counted? When ENOUGH ballots have been counted? When all recounts have been completed? When appeals/legal challenges have been exhausted?
Or perhaps all of that is irrelevant, and it's when the electoral college meets that matters, which was December 14 in 2020? (That one's actually reasonable, and Trump lost that one.)
Does Trump agree that he lost that one?
I don't believe he has issued any official statement about it. But also, it's the certification on January 6 that determines whether or not he (retroactively) lost on December 14. At any rate, I don't think the precise time one officially wins is something anyone has cared to adjudicate, since the inauguration (specifically the moment he completes swearing the oath of office) is generally what matters.
I can't tell whether he conceded the election, but on 1/7/2021 he conceded Biden would be inaugurated. https://www.cnn.com/2021/01/07/politics/trump-biden-us-capitol-electoral-college-insurrection/index.html
All of the above. Every President Elect has violated the law halfway through their first nomination.
No, only every one since Eisenhower.
I don't know where I originally found the link, but this guy Sam Jennings has the best take I've seen on Trump's victory. A few extracts:
> Which is not to say I'm excited for the coming chaos. Only a very certain, or very rash, person would prefer unbridled chaos to an anxious blandness. And yet we have no real choice in this. The people have chosen chaos. They’ve hit the big red button that says, “Go on, fuck it all up, what do we have to lose?” and now we’ll get to see exactly what we have to lose.
[...]
> It’s an awful picture, yet I can think of no other way to put it: the entire attitude post-election seems to me to be essentially post-ejaculatory. We’ve been anxiously invested and secretly turned on by a carnivalesque libidinal ritual, being hacked out by two very unsexy parties, and now we’re all emerging from a vaguely masturbatory haze to discover how sordid and sad the whole thing was.
https://samueljennings9.substack.com/p/please-shoot-the-messenger
> and now we’re all emerging from a vaguely masturbatory haze to discover how sordid and sad the whole thing was.
For some? Sure. For Trump's core base? Smells like projection. People like to pretend that Trumpism is just some inexplicable wave of hysteria. When in fact, the stage was set 30 years ago [0]. Keynes once said "When the facts change, I change my mind. What do you do, sir?" Well apparently, Jennings' go-to strat is to revert to epistemic helplessness. It's not as bad as full-on denial, I'll give him that.
Say it with me, everyone: "We're not at the end of history"
[0] https://www.astralcodexten.com/p/open-thread-353/comment/74547663
My favourite take was that Kamala Harris ran a campaign about ideas, and that non-college educated people voted as a bloc for the person whose campaign was about them.
I don’t think everyone who voted for Trump just wants to watch the world burn, as noble as that is.
Lots and lots of people who voted for Trump but want opposing things think Trump is going to be on their side. They can’t all be right! It will be interesting to see how it shakes out.
NYMag did the thing where they actually asked Trump voters why they voted for him, recommended read:
https://nymag.com/intelligencer/article/donald-trump-new-york-election-results-turning-red.html
Does that magazine by any chance make it possible for you as a subscriber to create a non-paywalled reading link to that specific article?
Use https://archive.ph/
Well I guess now's the time to link that NBC News story I saw about Kamala Harris raising $1.4 billion and blowing through all of it in four months. https://www.nbcnews.com/politics/kamala-harris/clashes-confusion-secrecy-consume-harris-campaign-finances-rcna179654?utm_source=pocket-newtab-en-us
We can watch what they did with it. https://www.youtube.com/watch?v=npxb6vytkZg
It's an unfounded claim to say Kamala Harris would have been anxious blandness and not unbridled chaos.
How is it unfounded ? It is very likely that Harris would have kept on with the current policies. It is also very likely that she wouldn't have tried to break any norms.
I keep seeing people trying to "both sides" this so hard but it always seems so dishonest. For example equating Hillary's "[Trump] knows he’s an illegitimate president" which is calling out shady voter suppression tactics and the fact that Trump did not win the popular vote with the staunch denial of the 2020 election result by Trump to this day and the organization of a (failed) plot to remain in power.
Are you people serious ? Do I really have to convince people in the comment section of a rationalist author that Kamala Harris was willing "to destroy the country" ?
"It is very likely that Harris would have kept on with the current policies. It is also very likely that she wouldn't have tried to break any norms."
I agree with this 100%. That said, I have a different interpretation: if you keep on with the current policies, and you don't break any norms, the country is going to be destroyed. And she's "willing" to let it happen. She doesn't think it *will* be destroyed, of course, but that's where we disagree.
Well at least we share the same reality but if you think keeping on the tracks of the previous administration is going to destroy the country you may as well define what destroy means to you because it seems far off the center of even the metaphorical meaning of this word.
https://www.lesswrong.com/posts/yCWPkLi8wJvewPbEp/the-noncentral-fallacy-the-worst-argument-in-the-world
Yes, they are; but no, you don't. I put it to you that there are better uses for your time than trying to interrupt people who are deep into their enjoyment of a gratifying fantasy that, while absurd, can never truly be debunked.
The Harris administration will always be a counterfactual. You can't even prove to *me*--a person who thinks it's broken-brained delusion to believe otherwise--that President Kamala Harris would have continued to be a bog standard bland Democratic politician, presiding over a bland (but diverse!) Democratic cabinet and pursuing vaguely center-left policy objectives, to the extent possible with a divided Congress, in the usual boring and bureaucratic manner for four to eight years. How do you expect to prove it to people who genuinely get a kick out of imagining her becoming Woke Hugo Chavez overnight?
"You can't even prove to *me*--a person who thinks it's broken-brained delusion to believe otherwise--that President Kamala Harris would have continued to be a bog standard bland Democratic politician"
Look at the record. She got a reputation as "Copmala" in California. Later on, when first running for the primary, she tacked hard left because the winds of change had shifted to a different position being considered more appealing to the public. That got her nowhere, so this time round she moved back to the centre and avoided explaining why she was repudiating her former positions.
That's because she doesn't *have* any positions. As Vice-President, she's best known for what? A media controversy over whether or not she was known as/appointed as "Border Czar"? Her campaign this time round ran on vibes.
If elected, she would have continued to be as bland as rice pudding and would just have gone along with whatever seemed to be the best position to give her the greatest chance of high approval ratings. She wouldn't have gone full socialist and she wouldn't have reverted back to tough on crime (legal weed and free loans for black men to transition from being drug dealers to being legal drug dealers/businessmen still has me face-palming here).
Yes but that's a good thing in a democracy. You want someone who will do what's required within the limit of the system. Always going the in the direction of the person in power's latest whim is not a good thing. It strains the system and may end up breaking it. I'll tank blank rice pudding instead of castor oil any day.
Well, she took a strong stand against the First Amendment:
https://www.washingtonexaminer.com/opinion/columnists/3142667/harris-banana-republic-free-speech/
>In 2019, Vice President Kamala Harris told CNN’s Jake Tapper that social media companies “are directly speaking to millions and millions of people without any level of oversight or regulation[censorship] and _it has to stop_.”
[emphasis added]
And, in addition to her words, she was a participant in the Biden/Harris administration, which used Facebook and pre-Musk Twitter as private proxies for unconstitutional censorship by their administration. So this is not just a hypothetical of what Trump _might_ do.
There's an argument that she was pro-censorship merely because that's what the polls suggested her voters wanted. If the polls said her voters were 60-40 in favor of tariffs on quinoa, Tapper would have been hearing her rail about how Peruvian agri-conglomerates "have put thousands of small American quinoa farms out of business without any remedy and _it has to stop_."
I took this as Deiseach's point: Harris lost appeal because she resembled a weathervane. Even her speech stance might have been poll-driven. (I think it's likely, and all it would have taken to flip her would be a minor BLM tweet storm.)
I heard she has a communist flag in her garage.
It occurs to me that you never actually /hear/ this sort of thing any more. It's too gauche to mudsling out loud, even at the end of a night's drinking. Mostly, now, you /read/ it in the more shouty bits of the internet.
I was making a joking reference to a 1973 ‘novelty song’, Uneasy Rider by The Charlie Daniels Band.
https://en.m.wikipedia.org/wiki/Uneasy_Rider
I’m running on gallows humor these days.
>It is very likely that Harris would have kept on with the current policies. It is also very likely that she wouldn't have tried to break any norms.
Based on what?
The money just changed hands; no wealth was lost.
Same with burglary.
...What's so horrible about that ad? That seems like a completely run-of-the-mill political ad. I thought you were going to link something calling Trump Hitler or something, but no, it's just making the obvious point that Trump is not aligned with the working class in any capacity.
Ah, it's that the guy is obviously An Actor playing the part of "regular guy who goes to football matches with his dad". Sure, a real Regular Guy would have been stiff as a plank in the part (if you've ever heard radio ads with Real People, you'll know what I mean). So it's plastic not real wood, while the tone of the ad is "we're real, he's not".
Harris' disdain for the common man bleeds from it. There's the prevailing sense that she thinks people are cows and they'll follow whoever looks most familiar. "He lives in a country club, that's not your herd! I'm your herd, with the haircut and the Sports!"
This is a twisted and uncharitable take. Do you really think Harris personally oversaw this ad ?
A lot of ads I see from the US country are like that : Interview someone that is supposedly just one of the viewer so that they identify and buy something. Nothing special about it.
No, I doubt Harris had more to do with it than "my campaign signs off on these". It's a terrible ad, but no worse than many political ads, so she can't be blamed for that.
But "Trump is a fake!" messaging loses out when the guy spouting it is clearly professional actor doing a role and he'll be dressed up as white collar PMC guy, possibly IT guy, for another ad instead of "I'm a guy what drives a SUV and work in a saw mill".
Contrast that with the garbage truck, and that's really Trump in a real garbage truck. It may be dumb, but it's funny and most importantly it's not faked up the same way.
Or rather, it's so OBVIOUSLY fake, like his McDonald's photo-op, that it has its own kind of authenticity.
Harris is in charge of the organisation that made this ad, her personal level of involvement doesn't matter. Being senior management means taking responsibility for the actions of the organisation that you built, and Harris was asking for the most senior-management job of all.
I'd never seen this ad before. It's not quite as terrible as the "I'm a man what eats carburettors for breakfast" one, but it oozes the same contempt for voters. "We have to figure out the things that each of the demographic groups on rows A through M of this spreadsheet likes, and make an ad filled with each of them, and say Donald Trump Bad. Make sure it doesn't mention Kamala Harris or any actual policies."
>Do you really think Harris personally oversaw this ad ?
It ends with "I'm Kamala Harris, and I approved this message", so yes, she personally oversaw it as much as she was going to personally oversee the presidency.
Sure
https://www.youtube.com/watch?v=V0pN1PpfIag
...I see your point. Still, I don't think it comes from a place of contempt. There is a genuine lack of understanding on why the working class loves Trump so much, fueled by blind celebrity worship...
I guess the reason doesn't matter though. The consequences for failure are all the same, and I have no sympathy for the left's coming death.
I guess it’s a Rorschach test.
I probably have as much blue collar cred as anyone here and I see no contempt.
Very interesting. Read into this (commercial) what you will.
The campaign funding reminds me of 2016 when Hillary's campaign was boasting of all the money it was raising and there were constant media pieces about "Trump has no donors, no money, his campaign is going to run out of steam soon".
Well, looks like the deeper pockets didn't win a second time! And Trump's campaign didn't need to spend money on publicity because everyone was throwing free publicity at it with both hands! From "The Republicans are weird" to every breathless "it's the Fourth Reich for sure this time!" social media/traditional media take, they honestly spent more time talking about Trump than their own candidates.
What was Walz gonna do in the White House? I have no idea, but I certainly know the Harris campaign position on JD Vance.
The difference this time round seems to be the begging letters/emails being sent out still, and in increasing numbers. The over-spend seems to be about $20 million, which is not terribly bad, but the reaction makes me wonder if there is more there we are not hearing about. Is the DNC growing a spine and demanding Harris personally take responsibility for paying it off or making sure it's paid off? Are the deep-pocket donors cajoled into backing the booting out of Biden now having buyer's regret and intimating they may not be putting their hands in their pockets the next time a Democratic candidate comes looking for a little help to protect democracy? Reading between the lines, there seems to be some indication of the usual bribery re: walking around money to get out the black vote, that may not sit well with the emphasis on transparency and openness and 'Kamala was the grassroots choice'. Also that consultants, hangers-on, and 'jobs for the boys' chomped a good bite out of the funds, but that's par for the course in politics.
And of course a ton of unforced errors:
https://nypost.com/2024/11/09/us-news/trump-trolls-debt-ridden-harris-camp-urges-gop-to-do-whatever-we-can-to-help-them-after-its-revealed-vp-owes-20m/
"Harris camp also spent a whopping six figures building a set for her appearance on the “Call Her Daddy” podcast, which attracted a fraction of the viewership of Trump’s appearance on the “Joe Rogan Experience.”
Yeah, that sure was worth the money they tossed down the sink, wasn't it? This kind of thing niggles at me - the campaign seemed to want a huge level of control over such interviews (not going to the studio where this podcast is usually held, and the same with what Joe Rogan has said about them - they wouldn't let her go to him, they insisted he had to go to them) which sounds like they didn't trust their own candidate enough to let her do an unscripted interview on strange territory. That doesn't bode well for "she is competent and capable".
There seems to be an air of desperation that I don't remember from Hillary's loss this time round.
Matt Shapiro had a (paywalled) post about against the grifters that I thought was somewhat relevant
https://www.marginallycompelling.com/p/against-the-grifters
I think I might disagree with him in that in many ways the grift is the point for a lot of Dems; I think my best steel-manned version of this "functioning as a political party that wins elections is secondary to its function as social club for the Professional Managerial Class and above, and throwing out large contracts for bad services is merely a cleverly-disguised welfare payment
To that last point - from my recollection, the atmosphere in November 2016 was complete shock - "how *could* this happen" overshadowing "how *did* this happen", with blame focused more on Americans failing to elect Hillary than her failing to appeal to them. The latter came later.
I've been deliberately avoiding Takes in all directions, subjects and temperatures the last few weeks but I really enjoyed this, thanks for sharing
A couple of thoughts. One, I don't think the people who voted in Trump are having any kind of second thoughts or "post-ejaculatory" worries. They seem pretty content with having won and are looking forward to seeing how far Trump can get with fixing their problems. I don't think many, maybe any, have delusions that everything will go their way, but they have a reason to hope that some will go their way. Second, and this is important for people who hate Trump to understand, Trump only exists because the people who support him have felt like they didn't get *any* of their preferred preferences since sometime in the 80s or 90s. NAFTA and greatly increased immigration have done huge damage to them economically and socially. If the choice is "more of the same" or "unbridled chaos" they pretty much have to pick chaos. It's the choice between some level of hope and death.
> If the choice is "more of the same" or "unbridled chaos" they pretty much have to pick chaos. It's the choice between some level of hope and death.
How can this be true for Trump voters while all the surveys show that liberals in general are less happy than conservatives?
It could be true if conservative living is just that much better than liberal/left. That's not a charitable option, but there are so many possible confounders that I don't know that we could ever really determine the reason(s).
My personal take is that rural living really is that much better for human happiness, and rural people are significantly more conservative. Less crowded, more likely you know your community members. Less chance of running into random jerks and criminals, even if the rate of criminality/jerkiness is the same or worse in rural areas. There's something to be said about the ability of one jerk to ruin your day, and the chances of running into one jerk a day is unbelievably higher in a city (unless the jerk is your family or a close neighbor or something).
There's two elements to it.
The first is that the conservative (small c) life script is genuinely better for human flourishing. Finish school, get a job, get married, have kids, go to church and engage with your community and extended family (all while living in the countryside or suburbs).
Compare with 'Finish school, do a postgraduate degree, move to the city and try to break into a competitive field, sleep around for several years, cohabit and then maybe have a child in a small apartment without getting married'.
All the elements of the first option are better for psychological health.
The second element is that the conservative worldview itself is better psychologically. Believing that you are responsible for your own fate, believing that the world is basically just, believing that you have duties to other people and to ideas bigger than yourself are all psychologically helpful.
Jonathan Haidt has done some interesting work on this. He refers to the modern leftist worldview as 'anti-CBT' for this reason.
This, basically.
I don't really agree with most of their policy preferences, but society has been running on "Certain people have gotten their way too long and don't get to have their way anymore" for so long that half of the members of "certain people" haven't actually seen anything go their way for their entire lives - up until Trump. Society has been on this train for about thirty years now.
("Certain people" is not "white people", or "men", or "Christian"; it's both more specific and more general than any of those things.)
I expect some people have a few minor qualms about the RFK nomination.
Yes, I think that the process of being disappointed by the politician you supported is usually a gradual process, not a rapid one analogous to post-nut clarity.
I think that Harris supporters might be having a quicker-than-usual realisation that she was shit, but that's only because they all thought she was shit four months ago and then suddenly turned on a dime the moment she became the only possible nominee.
>I think that the process of being disappointed by the politician you supported is usually a gradual process
Well, I viewed the contest as between Horrible and Worse (with Harris _narrowly_ getting the role for Worse), so I'll only _really_ be disappointed if Trump actually manages to stumble into a nuclear exchange (unlikely, but possible - but it was possible with Harris too).
Given that the only other choice was an idiot fascist, you gotta line up behind the lukewarm neolib. I mean, the last lukewarm neolib squeaked his way in.
Who knows though, maybe running a centrist campaign and getting blown the fuck out will knock something loose in the D's hydrocephalitic party apparatus. One can hope.
I got the impression Harris was trying to project centrist and people didn't buy it. After four months of Harris being in the spotlight, I still have little idea of who she really is. The best glimpse I got was during the debate, for which she had a month to prepare, and all I saw was preparation.
>who she really is.
Evasive? :-) (That was the main impression I took from the debate, and from the Bash interview.)
Projection is the same as reality in this case.
People who personally hate trump and generally don't like the paleocon style politics of his hangers on didn't get fired up by:
"No socialised medicine, Expand Drilling, Support Israel to the hilt, I Love Rich People, Here I am standing next to a fucking Cheyne" style messaging, and your actual conservatives and centrists (embarrassed conservates) were never going to vote for a black woman running against their favorite idiot old man on the back of "Look, I know everything is more expensive, rates of savings are the lowest they been for decades, you are one broken leg away from missing rent and being homeless, but line go up!"
"I still have little idea of who she really is... all I saw was preparation"
I never understood this argument, bc it always seem like any criticism you can levy at Kamala would be one thousand times worse with Trump. No, I would never prefer an honest Hitler to a passive-aggressive faux-kind co-worker.
Yes, she prepared for a debate, sure, she isn't being super genuine in her policies, but is that really worse than Trump? Someone who didn't plan at all for the debate, didn't have any policies proposals ready, and lied through his teeth? Trump is a genuinely stupid man, who does not understand much about how the government functions. How is he any better?
People wonder about how a civil war in the US might break out. How plausible does the following scenario sound?
1. Trump purges the army of commanders not considered loyal to him (under the guise of some other reason)
2. Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
3. 2028 rolls around and a Democrat is elected POTUS. He tells the army to withdraw from the cities but meanwhile Trump tells them not to and at least one of the armies obeys Trump not the new POTUS. (Throughout the period after martial law is declared, The Left stays angry about the situation, with violence breaking out occasionally, army personnel sometimes getting shot by The Resistance.)
4. Some generals stay loyal to Trump while other generals are loyal to the POTUS.
5. Civil war.
How plausible does something like that sound?
I think waiting until 2028 would be a bad play for him. It would give the opposition time to find a figurehead to coalesce around. Shutting down the opposition should happen before then, optimally during or a bit after the midterms. Of course, he would need to artificially inflame tensions domestically; possible options for that are doing something blatantly unconstitutional that is still popular with his supporters, the deportation program getting revealed to have caused deaths or other human rights abuses, or simply enabling Israel to finish the job in Gaza.
wrt my expected response to this question:
“Unleash the hounds!” C. Montgomery Burns weakly shouted.
Highly implausible.
Very unlikely. Every stop of your process seems individually unlikely, with a combined effect of all but impossible.
1) Gutting enough of the military to make it specifically loyal to Trump would completely eviscerate the military. Rank and file people may love Trump, but not separate from their love of the US. The commanders don't particularly love Trump.
2) Limited martial law is possible, but not multi-year. The federal government calling for martial law would likely result in local jurisdictions doing everything in their power to fix a problem before federal troops could even get there, and then quickly getting them out. Martial law is bad for everyone involved and would be incredibly unpopular, especially after an immediate need was handled.
3) This seems to presuppose that Trump doesn't leave office? I'm not sure what you have in mind, but once we have a new president who can give lawful orders to stand down, then we don't have Trump in office to give counter orders. If Trump would fight against leaving office then the details of his argument and who supports him matter. I don't see that as being very likely, either.
4) If Trump could pull off the complete gutting of the military, I guess some of the new generals could potentially support Trump. But they would almost certainly be very bad generals, likely incompetent at their jobs and unliked by their troops. If they were qualified people who rose in the ranks after their bosses were removed, then why follow an illegitimate Trump?
>1. Trump purges the army of commanders not considered loyal to him (under the guise of some other reason)
Seems definitely possible. I'd say at least 50% chance this happens anyways.
>2. Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
Very unlikely. Maybe 1-3% chance IMO. There are a lot of other steps before that: declaring national emergency, militarizing the police, sending in the national guard, federalizing the state national guard, repealing or ignoring all the laws that prevent him from doing this...etc. etc. Public sentiment would be low. Congress would likely act against it. Supreme Court is, contrary to popular opinion, not a bunch of Trump loyalists (they're FedSoc which is very much not a MAGA organization!) and would probably act as well. States would likely refuse to cooperate.
>3. 2028 rolls around and a Democrat is elected POTUS. He tells the army to withdraw from the cities but meanwhile Trump tells them not to and at least one of the armies obeys Trump not the new POTUS. (Throughout the period after martial law is declared, The Left stays angry about the situation, with violence breaking out occasionally, army personnel sometimes getting shot by The Resistance.)
less than 1% chance. Highly unlikely
>4. Some generals stay loyal to Trump while other generals are loyal to the POTUS.
less than 1% chance. Highly unlikely.
>5. Civil War
less than 1% chance. Highly unlikely.
I can think of like 10 different possible paths off the top of my head to civil war that are more likely than this. Most involve some kind of state secession crisis or medium-intensity insurgency, rather than a president trying to become king.
> There are a lot of other steps before that:
> militarizing the police
How would the police be different after this process than they are now?
My guess would be to sequester them from the general populace, official purpose being "special training". Some of them are selected for special teams, given special gear and instructions for using it.
Generally, any government that wants to oppress its own people gets that going by having a police force that doesn't socialize with those people. Very, very unlikely in the US as a result. The closest we tend to get is hiring police from one part of town to police another part of town.
On "1" there's a big gap between "personally loyal to Trump" and "actively disloyal to Trump". I might expect him to fire a few of the actively disloyal ones, but the vast majority of officers are in the middle, willing to obey all legal orders.
Good clarification. That's more of what I had in mind.
> Martial Law is declared due to major protests or riots over something (something like the George Floyd riots), and the army is brought in. The US Army stays visibly positioned in several major US cities.
Cities are wildly leftwing, and quite well armed in practice
Riots happen because they are allowed to happen, if trump is threatening to end them using the national guard and it would be mildly embarrassing, the riots will stop.
> How plausible does something like that sound?
The administrative state donates like 98% of the money they spend on elections to democrats; the idea trump finds people to staff the current state with lackeys is just wrong. Even if he purges(fingers crossed), how will he find the manpower of people trained in the machine without dipping into those who are merely loyal to the paying the bills.
If trump changes the character of the american empire, the american empire is getting much smaller at which point it will be more of a states rights issue, unless you believe trump will control Californians governor.
----
I dont see how a rightwing-Establishment backed civil war happens, its the leftwing using the Establishment vs the right wing rural factions(with you know, guns, food and land; farmers protests are very powerful for a reason) thats ever been on the table, cities vs rural is the fundamental divide in politics and always will be, food needs land, high speed trade need dense population, cities rarely see the farms but are wildly dependent on them.
And the country has spoken, the outsider is invited in to rebalence the scales, he will be surrounded by a bunch of leftwing insiders looking to block him at every turn; this is good for peace(assuming he survives assassination attempts)
"How plausible does something like that sound?"
About the same as winning the Powerball.
Incredibly unlikely. Trump purging enough people from DoD to where they would support long-term suspension of Posse Comitatus to implement Martial Law is one where the Army wouldn't be competent enough to do it anyway.
I don't know if it's been said on here before (and I'd love to hear @Deiseach's perspective on this), but elsewhere it's been noted how fascinating it is that people's mental model of an American Civil War is something that looks a lot like the Spanish Civil War and not "Na Trioblóidí, but bigger". Because I think we're a lot closer to the latter than I think we would like to believe, and I think the former is much harder to make happen than I think people expect, to say nothing of "Nicaragua but with more paramilitary death squads".
The divisions in the US are nothing like Northern Ireland. You asked Deisheach who is, Irish. I’m British with family in Northern Ireland - so with an understanding of history of the place.
These ethno religious civil wars are not what is happening or possible in the US. An ideological civil war is unlikely either, despite the caterwauling about Trump, he is not a fascist and the opposition are not communists either. In the Spanish civil war they were, literally that. That, and anarchists. Democracy was a young political idea then and had little credibility.
In the US it is part of the fabric, however flawed that fabric, for centuries.
I did worry slightly about the US creating, out of nowhere, inter ethnic conflict at the height of the anti white supremacy kerfuffle a few years back, those worries have entirely dissipated in this election. It looks like Hispanic males in particular either like a bit of white supremacism or don’t believe it.
Being old enough to be around during the 70s and just old enough then to be somewhat aware of what was going on in America, I remain astounded you didn't have your own version of The Troubles. You really got lucky and zigged where other countries would have zagged. You had your own amateur terrorist militias robbing banks and performing bombings, you had political theorists pushing revolution, and there is of course the race question.
Seems like your Founding Fathers got something right when setting up the state!
Probably helped a lot that our most dedicated and extreme revolutionaries could always defect and go back to cushy jobs thanks to being rich kids. No need to be a diehard martyr, you can just become an education professor!
https://www.amazon.com/stores/William-Ayers/author/B001JPBUT0?ref=ap_rdr&isDramIntegrated=true&shoppingPortalEnabled=true
Try iodine. If that doesn't help, organic food. Sulfonylureas are used as preharvest desiccants in conventional agriculture. They're also used as diabetes drugs, because they make your pancreas squirt out insulin. Long-term, this can lead to an insulin resistance phenotype, and hypertriglyceridemia can follow on from that.
I might follow up on the iodine, can't hurt!
Im an anti-reductionist; I believe logical positivism was proven wrong a century ago with gobel incompleteness but 99% of people just let it continue to effect their thinking with things like the simulation theory(what if the 3 body problem is unsolvable computationally, there wont be an infite chain of universes being simulated as each simulation wastes more resources)
This argument doesn't seem to work for most people, its to foundational, and maybe its to big a epistemic switching cost? So new thought experiment, the color blind prisoner:
You are colorblind and at the mercy of a warren who is a full-sighted supremacist, he is running "games" to find who to free and who to kill, you being infinitely smart deduce his methods of randomly assigning colors, the flaws of the coins so you have 50.0001% at predicting his coin flips etc. Do you really think you survive? Or do you fundamentally lack the right sensor? Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
It is currently believed in this community that ai will deduce physics of the real world from video: https://www.youtube.com/watch?v=fVN_5xsMDdg why? How do you discover dna without a microscope? How do you discover quantum physics without the double silt? How do you discover trustworthy faces without human feedback?
Critical data exists, and is an "unknown unknown" until stumbled upon; you must poke and prod reality you cant merely watch and learn everything.
> It is currently believed in this community that ai will deduce physics of the real world from video
What?? Aren't physics textbooks already included in the training data for every LLM?
I'm not 100% confident I understand. But if you're saying "at some point, you need to stop armchairing and use your eyeballs", that's a theme Yudkowsy has definitely covered. In fact, that's the entire point of his obsession with Bayes' Theorem. More specifically, there's a post that comes to mind where he talks about how syllogisms only tell you about possible worlds, and you need evidence to narrow things down further. I'll edit if i can find it.
edit: I think it was (https://www.lesswrong.com/posts/bcM5ft8jvsffsZZ4Y/the-parable-of-hemlock).
I dont know why 3 of yous have brought up yudkowsy, do all`y`alls not know the logical positivists?
>>> It is currently believed in this community that ai will deduce physics of the real world from video: https://www.youtube.com/watch?v=fVN_5xsMDdg why?
> I dont know why 3 of yous have brought up yudkowsy
It's time to recite the lore of yore.
Before Scott blogged on Astral Codex Ten (ACX), he had another blog called SlateStarCodex (SSC). Scott deleted SSC for a few months(?) in retaliation toward the New York Times (NYT). He gave an interview to the NYT about COVID-19, but the NYT used his real name (n.b. "Scott Alexander" is a pen-name) and Scott preferred not to self-dox for various reasons. Before SSC, he blogged over at Lesswrong (LW) under the handle "Yvain". LW is a community blog, and was the epicenter of the diaspora of Bay Area Rationalists. LW was founded by Elizer Yudkowsky (EY), who was obsessed with AI x-risk and Bayes's Theorem, among other topics. Before LW, Scott had a livejournal. I forget his handle. Scott and EY were the two most prolific writers on LW, I think.
The "ai will deduce physics of the real world from video" idea comes directly from EY's posts on LW about Solomonoff Induction (SI), AIXI, etc. More specifically, the Rational Animations video is a roughly paraphrased narration of EY's "That Alien Message" [0].
> Riemann invented his geometries before Einstein had a use for them; the physics of our universe is not that complicated in an absolute sense. A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a hypothesis — perhaps not the dominant hypothesis, compared to Newtonian mechanics, but still a hypothesis under direct consideration — by the time it had seen the third frame of a falling apple. It might guess it from the first frame, if it saw the statics of a bent blade of grass.
In all likelihood, the three of us just presumed you were familiar with the lore without looking at the video, and were knowingly engaging with EY's arguments. Now that I've watched the video, I don't think it did a great job of explaining what the actual point of the story was. EY's claim isn't that AIXI can deduce all of physics from a few video frames with absolute certainty, but that it can at least *brainstorm* lots of plausible theories, based on the assumption that the physics of reality aren't pathologically obtuse, and that AIXI subjectively experiences reality millions of times faster than humans.
(N.b. I have my own opinion of EY's argument. But first, it's important to establish what the argument is, and where it comes from.)
(Edit: for additional context, EY claimed that Bayes' Theorem is a generalization of the scientific method. Bayes' Theorem is useful when you're poking, prodding, and A/B testing when you're allowed to do that. But unlike science, it's *also* useful for times when all you can do is passively observe, like in the case of astronomy. Maybe this helps clarify WindUponWaves' comment.)
[0] https://www.lesswrong.com/posts/5wMcKNAwB6X4mp9og/that-alien-message
> (what if the 3 body problem is unsolvable computationally, there wont be an infite chain of universes being simulated as each simulation wastes more resources)
I don't think that's what people mean when they say the 3-body problem is unsolvable.
AFAIK, the 3-body problem is *chaotic*. "chaotic" means if you make a prediction about a system's output, it can wildly diverge from reality because the measurement-error snowballs out of control. You can certainly simulate it, though. E.g. a pinball machine is "chaotic" in the sense that a tiny change in the angle of its initial trajectory can lead to the ball flying in opposite directions off its next rebound. If you were to try to "solve" the path "analytically" by matching it to a one-size-fits-all shape (like say... an ellipse), you're going to be disappointed. But if it were unsimulable, pinball videogames wouldn't exist.
Those are the same thing.
If you cant simplify the 3 body problem then simulating an electron with say 3 64 bit floats, means you used on the cheap end you used 192 electrons to simulate 1; this stops infinite regress.
Oh, that sounds like an observation about the map necessarily being smaller than the territory (else it's not a map, it's a copy). In which case, mentioning the 3-body problem obscures more than clarifies.
I mean, I kinda see what you're getting at, from your examples. But I suspect you're confused about certain concepts.
A) Logical Positivism embraces empiricism.
B) Bay Area Rationalism embraces empiricism.
C) 17th Century Rationalism *shuns* empiricism. (Many, such as TheAncientGeek, have complained that EY's name for his community was *very* poorly chosen.)
> I believe we are seeing glimpses of a non-reductionist biology and we will *finally* discard to notion that any system evolution creates only does 1 thing at a time with cleanly defined boarders.
D) Uh... I don't think that's what "reductionism" means? Nor is that representative of what biologists believe? Where are you getting this from.
> You are colorblind and at the mercy of a warren who is a full-sighted supremacist, he is running "games" to find who to free and who to kill, you being infinitely smart deduce his methods of randomly assigning colors, the flaws of the coins so you have 50.0001% at predicting his coin flips etc. Do you really think you survive? Or do you fundamentally lack the right sensor? Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
btw, this sounds like the Multi-Armed Bandit Problem, which highlights the "exploration vs exploitation" trade-off. I.e. "is it better to min-max a particular slot-machine, when a neighboring slot-machine might practically print money for you?" I remember there was some discussion in LW's Geometric Rationality Sequence [0] (not EY) regarding the Bandit Problem and Thompson Sampling. In case you're interested in reading or participating.
[0] https://www.lesswrong.com/s/4hmf7rdfuXDJkxhfg
Yudkowsky is just describing history. He's not even speculating, he's just describing what's already happened, because his claim is a lot more modest than people think.
As in, what he actually said was,
"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a *hypothesis*..."
Not that it would prove it, or deduce that it must be true. Just come up with it, in a "throw shit at the wall and see what sticks" way, like what particle physicists and string theorists do while waiting for the next supercollider or gravitational wave detector to be built. If you read or listen to the original context, it almost sounds like he's describing a fandom coming up with wild ass guesses about what the rest of the series holds, because they're impatient, this is good fun, and just by chance *one* of them is probably going to be right if they cast a wide enough net.
(In fact, my college roommates and I did exactly that, about the physics of a hypothetical universe -- The Prestige, regarding Tesla's cloning device, if you're curious. One of the best times of my life...)
Anyways, Yudkowsky isn't even speculating. He's just describing what's already happened in history. The classic example would be the mathematics of ellipses -- in short, they were invented by the Ancient Greeks for such sterling uses as "describing the funny shape the shadow of a Sundial tip makes as the Sun goes round" and "trying to do something that's literally impossible, a fool's errand" (doubling the cube, a.k.a. squaring the circle, a.k.a. construct a cube that's precisely twice the volume of another cube, using only rational numbers, despite the fact that the cube root of 2 is irrational. But the Greeks didn't know that. What they did know was that ellipses were useless for practical matters like astronomy, because the orbits of the planets are *obviously* perfect circles, as befits their august position in the heavens).
Then, 1800 years later... those ellipses turn out to perfectly describe the orbits of the planets, once you gather more data by inventing telescopes. In fact, they're necessary for things like "Heliocentrism" to work, because otherwise Galileo-style Heliocentrism is *worse* at predicting the orbits of the planets than the Ptolemic Geocentric model, and no one will take Heliocentrism seriously if it's simply worse at the job.
But instead of the scientific community needing a lifetime of work to try out all possible orbit shapes and the math necessary to make each one work... Kepler reads an old book and realizes the job has already been done for them. The lifetime of mathematical work has already been done for them, ahead of time, by someone they've never met. Menaechmus the Ancient Greek mathematician had never looked at a planet through a telescope, and might not have ever actually *seen* a planet in the night sky at all if his eyesight was bad enough... but his work describes the orbits of the planets nonetheless. Not bad for someone who got even less than 3 frames of video!
It was all just dumb luck of course... but serendipity favors the prepared mind, and it turns out you can make *lots* of preparations if you've got nothing else to do while waiting for more data to come in. Menaechmus spent his idle time describing useless ellipses rather than perfect circles; Kepler spent his reading old books while waiting for his colleagues to report back on their telescope studies; and the two made a powerful team despite how stupid and ridiculous it all would have sounded at the time. "Oh, you 'theorize' the orbits of the planets *aren't* perfect circles, but instead the useless shape described by some dead guy? What's next, you *theorize* my dick into your mouth?"
This is hardly the only example, of course. Others include:
• Negative numbers (theorized about *way* before they were accepted as a useful and accurate description of anything in reality)
• Imaginary numbers (same)
• Topology (was the butt of jokes for generations; in other news, the 2016 Nobel Prize in Physics was for Topological Physics and its applications in quantum mechanics)
• Number Theory (G.H. Hardy, a British mathematician, felt so bad about this he wrote "A Mathematician's Apology" about why it was worth doing pure math like Number Theory when it was so obviously useless. His most famous defence was that, since it was so useless, at least it was pure of heart and untouched by war. That was written in 1940. Within a few years, Number Theory turned out to be key to cracking the Enigma, and cryptography more generally.
His other famous example was Special Relativity, of the "mass to energy" kind -- to him, Einstein's work was clearly useless for anything military like bombs.)
• Fast Fourier Transforms/FFTs (invented by Gauss, who thought it was useless; modern applications include detecting nuclear bomb tests through seismic waves, the MP3 file format, and indeed basically anything involving waves)
But the biggest example here would have to be... General Relativity itself. Or at least the mathematics behind it, for describing a Non-Euclidean universe of warped spacetime. By all sense and logic, Einstein should have been starting from zero when trying to describe how the universe isn't actually Euclidean, because who would waste their time investigating this possibility before him when there was precisely *zero* evidence of a Non-Euclidean universe, and *zero* evidence of how it would work? Bolyai, Lobachevski, and Riemann, that's who. They invented tools like Hyperbolic Geometry, Manifolds, the Riemann Metric, Curvature Tensors, and more, so Einstein didn't have to.
Why?
Because they enjoyed it, more or less. Their brains were too active to just be satisfied with the reality in front of them. Turns out they were actually describing a deeper reality all along, even if they didn't know it and were just shitposting at the time. But sometimes, throwing shit at the wall does pay off -- in this case, in *spades*.
So to summarize... that cute story about watching aliens, is actually just summarizing history in a digestable way. What do smart people do when they have nothing to do? They shitpost. They shitpost in such prodigious quantities, about the shape sundial shadows make or whatever, that sometimes they hit gold just by pure dumb luck. But that wasn't even the point, they simply can't help themselves. The allure of shitposting is just too strong. (And hey, sometimes it *does* pay off, 'specially when you've got nothing but time). Why should AI be any different?
Im not *only* talking about yudkowsky, if anything rationalists maybe a bit better at being fishes aware of water
> [humans are] "hooked up to a webcam"
> already happened in history
Your embodied in the world and some guy did just need to drop heavy objects to discover the greeks were wrong about heavy objects falling faster.
I just disagree that you only a meat computer encased in bone sphere, your your entire body; I understand thats gospel here and in the general society. But this is a new idea thats jumping to conclusions while ignoring data such as the possibly that you can treat depression with fecal transplants.
I have a blog post suggesting something even more radical and there michael levins work; but to simplify I believe we are seeing glimpses of a non-reductionist biology and we will *finally* discard to notion that any system evolution creates only does 1 thing at a time with cleanly defined boarders.
> no one will take Heliocentrism seriously if it's simply worse at the job.
People did and I believe I wouldve; heliocenterism was worse at predicting the data for quite a while and pushed by a socail outcast
> Imaginary numbers
> Negative numbers
> but serendipity favors the prepared mind, and it turns out you can make *lots* of preparations if you've got nothing else to do while waiting for more data to come in.
Thats not how I understand that history, and I view this as part of nonreductionism and would suggest the next step is surreal numbers; given a formal system with a unsolvable problem, label the simplest error case as a new primitive, extend the system.
A different way to describe negative numbers is letting the real world rewrite the axoims; animals may have a primitive counting circuit and old languages had no grammar to express negative numbers. Debt is a force of nature that humans took a long time getting a handle on.
"this statement is false" just say "tralse"(logical positivists spent decades trying to argue that such statements shouldn't even exist, before being proven wrong)
Theres the music theory thing of fifths being unresovable.
etc.
Formal systems will not survive contact with reality, this is simply an established fact; label the edge cases as new primitives and see what happens. This is not what the current culture promotes: negatives, imaginary, surreal numbers, follow this pattern but were done slowly and by rare individuals breaking rules.
Id advocate for doing the same with sleeping beauty problem and seeing what comes out of finding some new primitive for Bayes formula.
I think we're talking past each other, because it sounds like... you're agreeing with the story? That people can imagine things they have no evidence for, whether that be General Relativity or imaginary numbers or whatever. Which means they can stockpile ideas like tools in a toolbox, or keys on a key ring, and use them to very quickly open new and unexpected doors/solve never-before-seen challenges, when they *do* finally get the evidence, instead of having to slowly brute force them down. You're... agreeing with the story, not disagreeing.
I.e. the theme of the story can be summed up as something like, "The necessity of empirical evidence, and how to be efficient with it when you don't have a lot." The thrust of what you're saying seems to be "The necessity of empirical evidence, and how you're stuck if you don't have enough of it". Those... are 2 different ways of saying almost the same thing? I don’t see what the disagreement is. Both are ways of saying, "Empirical data is extremely valuable" -- it's just that one describes at so valuable you should be efficient with it, while the other describes it as so valuable you should get more of it. Those are 2 messages that work together, not fight, right?
It's worth being precise about claims that we're roasting. The original claim was that
"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a hypothesis—perhaps not the dominant hypothesis, compared to Newtonian mechanics, but still a hypothesis under direct consideration—by the time it had seen the third frame of a falling apple. "
This is basically just a claim about bit lengths. The entire text of Einstein's "RELATIVITY THE SPECIAL AND GENERAL THEORY" is 200 kilobytes, so that's a _very_ loose upper bound on the size of the turing machine that implements it. 3 frames of video is 3 * 3 * 640 * 480 * 2 = 5 megabytes. AIXI is definitely looking in to that. AIXI directly considers a remarkable number of hypotheses.
> It's worth being precise about claims that we're roasting.
Its not really even claims, its the implicit assumptions, such that precision is highly epistemological valuable
> The entire text of Einstein's "RELATIVITY THE SPECIAL AND GENERAL THEORY" is 200 kilobytes, so that's a _very_ loose upper bound on the size of the turing machine that implements it.
Strong disagree, you must consider the decoder program when talking about thoerical compression(the wikipedia compression contest for example has this rule), so add in several physics text books, cannon english, and human dna minium; im also inclined to believe in bloodline memory and the microbiome being relevant.
The message of "I need help" is not 9 bits long when communicated with an alien, "sos" is *heavily* reliant on precomputed context.
Lets operationalize this, what is the shortest turing machine that simulates the merger of three black holes? Does that seem like a reasonable way to phrase it, and if so are you actually confident that I couldn't produce such a turing machine under 200 kb in an afternoon of work by bodging together stuff I find on the internet? I did a brief try before getting too out of line posting wild claims, and I got as far as running two merging black holes in ~70 kb of C and so it's going to come down to how efficiently once can do C code to Turing Machine transpilation
I dont understand what that program would look like, or what its suppose to prove given blackholes are where the math breaks down and in theory "destory infomation"
As a public service I’ll add this:
bodging:
The origins of the term are obscure. There is no known etymology of the modern term bodger that refers to skilled woodworkers. It first appears c. 1910,[14] and only applied to a few dozen turners around High Wycombe, Buckinghamshire. The Oxford English Dictionary Supplement of 1972 has two definitions for bodger, one is a local dialect word from Buckinghamshire, for chair leg turner. The other is Australian slang for bad workmanship.[1] The etymology of the bodger and botcher (poor workmanship) are well recorded from Shakespeare onwards, and now the two terms are synonymous
New to my vocab. Thanks Hastings.
Actually I had to look up transpilation too.
I once converted some C++ code for a router for a nuclear power plant control system (a specialized network router) to functionally equivalent C# code.
Not sure if that would count.
But anyway, one day, two new words, I’m happy.
>Im an anti-reductionist; I believe logical positivism was proven wrong a century ago with gobel incompleteness
What's the relationship supposed to be? LP. Isn't the only support for reductionism. Did you mean Gödel ?
You can tell how well you are doing by how well you can predict : If you can predict most things you are either missing information, or It is not important.
>warren
?
> What's the relationship supposed to be? LP. Isn't the only support for reductionism.
The logical positivists seem to be the extremists in "truth is devised from gathering correct axioms and building up". Which is to say "truth is math" with more steps. Poppers falsification is a weak compromise.
I think darwin would be discarded under the modern bureaucratic science, Begging for mechanisms he never did provide (or couldve, biology is dense with chaotic systems) and when Mendel provided *partial* answers, he may have correctly predicted they were partial answers(tho we will never know, he never commented on Mendels work)
> You can tell how well you are doing by how well you can predict : If you can predict most things you are either missing information, or It is not important.
Its an unknown unknown if you assign an accurate upper bound on predictive power given no access to competitor gamblers.
Given a human market on guessing quantum events, perhaps we all tie for centuries on the current math; maybe some aliens have found an experiment that exposes a lower level and they know a 5 minute experiment to show us the key to the lower level. Or maybe we are correct and reductionism is very helpful for physics, I still dont think reductionism is correct for biology, people move on after getting partial answers "whats the purpose of breast milk" and attempted to make artificial milk before even learning about the immune system, much less our current bafflement about rapid directed evolution by an unconscious system but we do know immunity passes from mother to child.
Well, no, the LPs were much more empiricist than rationalist.
I'm still not grokking what you mean by reductionist.
>It is currently believed in this community that ai will deduce physics of the real world from video: https://www.youtube.com/watch?v=fVN_5xsMDdg why? How do you discover dna without a microscope? How do you discover quantum physics without the double silt?
Agreed (albeit quantum mechanics started from blackbody radiation and the hydrogen spectrum). We build instruments and run experiments for a _reason_.
>Is infinite detail of the irrelevant data worth 1 bit of the relevant data?
Agreed that it is not.
I think that's a mischaracterization of what the story is actually saying. I wrote a comment replying to Monkyyyhl here (https://www.astralcodexten.com/p/open-thread-356/comment/78027661), but in short, I think it's describing shitposting, of the kind familiar to 4Chan. Not gigabrained "We don't *need* data" 4D chess moves -- just garden variety "wild ass guesses", and the surprising ways that can actually pay off, just by pure dumb luck. After all, that's what already happened in history! Including with the story of General Relativity itself. (It’s just that this is an analogy the story didn't realize would help explain things.)
>Not gigabrained "We don't _need_ data" 4D chess moves -- just garden variety "wild ass guesses", and the surprising ways that can actually pay off, just by pure dumb luck.
Many Thanks! Ok, I would characterize some of those examples slightly differently.
>Einstein should have been starting from zero when trying to describe how the universe isn't actually Euclidean, because who would waste their time investigating this possibility before him when there was precisely zero evidence of a Non-Euclidean universe, and zero evidence of how it would work? Bolyai, Lobachevski, and Riemann, that's who. They invented tools like Hyperbolic Geometry, Manifolds, the Riemann Metric, Curvature Tensors, and more, so Einstein didn't have to.
Another way of looking at Non-Euclidean geometry is that mathematicians _routinely_ look at mathematical systems and see "Is there a more general version of this? Can we drop one of the axioms and still have interesting proofs?" Group theory can be looked at that way. Topology can be looked at that way.
And it isn't all that uncommon to find that some part of physics, which we'd thought was well described by a very restrictive model, actually deviates from that model - and the generalized model that the mathematicians came up with turns out to _still_ hold, including the newly found deviations from the old model.
In other words, the "unreasonable effectiveness" of mathematics looks a bit less unreasonable when considering that a substantial part of work in mathematics is looking for the answer to "Is there a more general version of this?".
Now, I'm _still_ skeptical about
>"A Bayesian superintelligence, hooked up to a webcam, would invent General Relativity as a _hypothesis..._"
A superintelligence, hooked up to a webcam, would have trouble noticing that the speed of _sound_ is finite, let alone the speed of light, let alone the speed of gravity. I don't think Yudkowsky has really realized just _how_ impoverished the evidence from a webcam looking at an ordinary scene _is_. For most scenes, all of the objects in view won't have more than invisible trace amounts of most of the periodic table. Most scenes will have neither fire nor ice. Most scenes will have no visible chemical reactions. Most scenes won't have enough inertial effects to motivate changing from Aristotelian physics to Newtonian physics.
I'm curious what experiment supremacists believe is happening in physics when Feynman diagrams are invented, or Einstein calls his shot with General Relativity, or we figure out the composition of stars from their spectra. Like, is it that theories can be right be sheer coincidence with very little supporting evidence? Or that they still had lots of evidence? But if it's true that they had a lot of evidence, just interpreted correctly, aren't we right back at where rationalists are?
Like, the world view *as stated* I believe would have said that what Newton did with his laws and orbital mechanics is impossible, that we could have made transistors faster *without* condensed matter physics, or that particle physics during its heyday was impossible.
What am I not understanding about the worldview? Or alternatively, what's wrong with my interpretation of those events?
Many Thanks! I'm confused about what you are asking or saying. Let me try to articulate my perspective on one of the cases you cite.
In the case of deducing composition from spectra, there was indeed a _lot_ of experimental evidence behind this. Lots and lots of observations of spectra of light from materials of known composition - sparks, gas tubes, flames, etc. .
And my main point above shows up here as the need for either a prism or a diffraction grating to see the spectra at all. One cannot just point a camera at a random scene and notice spectral lines. At best, one can e.g. notice the yellow glow of sodium in a flame, but finding the wavelengths of those lines, distinguishing from other nearby lines from other materials, that all requires specialized equipment.
None of this is to denigrate the role of theory. E.g. one needs the theory of energy levels and the transitions between them to turn information about spectral lines into information about the underlying energy levels, and then to tie that to thermodynamic equilibria to understand the temperature dependence of the intensities of the lines - which finally ties it back to stellar composition.
But I'm still confused about whether anything in this conflicts with or confirms your views?
> Lots and lots of observations of spectra of light from materials of known composition - sparks, gas tubes, flames, etc.
Yes, but crucially, none of that experimental evidence was from an actual star. How do you distinguish "you can't discover dna without microscopes" with "you can't deduce the composition of stars without going to one"? Because at the point where you are making statements like "we know burning sodium emits so and so" you *already* needed to have theories based on large amounts of extrapolation about electron ground states and what burning theoretically does. In fact, for something like general relativity I'm pretty sure we had "the evidence" that Einstein used for GR for around 50 years before he came up with it. If experimental evidence is the only thing that matters, why the lag? If reasoning was so important to connecting the evidence to theory, whence anti rationality?
I think by the time you admit "none of this is denigrating the role of theory", it's not clear to me you have anything to defend: someone who thinks they are much better than you at interpreting evidence can just claim that their theory is better, and then what does your philosophy do then? Yet there are strong statements here about how being pro experimental evidence *heavily bounds* the human ability. I just don't see this claim being able to make advance predictions rather than post hoc cope.
Many Thanks!
>Yes, but crucially, none of that experimental evidence was from an actual star.
True! The best we can do is admittedly a theoretical interpretation of what we've seen from light sources where we _do_ have access to their composition. ( Recently, also with information about the composition of the solar wind. https://agupubs.onlinelibrary.wiley.com/doi/abs/10.1029/1999JA000358 but that was decades after the original analysis of the solar spectrum and doesn't give us a check on the spectra of any other stars. )
>How do you distinguish "you can't discover dna without microscopes" with "you can't deduce the composition of stars without going to one"?
That's a good point. It's messy. For any question, we always have partial evidence, with varying degrees of indirectness between the evidence and the question we want to answer. Yes, spending effort on thinking through the implications of candidate models is very important.
My only point is that that computational/conceptual effort isn't the _only_ thing that is important. Instrumentation is really, really crucial too. There are many facets of reality which are just too minor under ordinary conditions, too swamped in noise, to see them without setting up the special conditions of e.g. a spectroscope.
>In fact, for something like general relativity I'm pretty sure we had "the evidence" that Einstein used for GR for around 50 years before he came up with it.
For the precession of the perihelion of Mercury, yes, for the deflection of starlight passing close to the Sun (yes, that was a test _after_ Einstein formulated GTR), no.
There are certainly cases where a conceptual snag is the limiting factor. Personally, I find it weird that mathematicians _still_ haven't been able to construct _a_ theory of quantum gravity nearly a century after Dirac extended quantum mechanics to be consistent with special relativity, and more than a century since Einstein formulated GTR. Formulating _a_ theory with the right limiting behavior to match the two existing theories is a conceptual/mathematical problem, not one of instrumentation - though _testing_ it would be an instrumentation problem.
>Yet there are strong statements here about how being pro experimental evidence _heavily bounds_ the human ability. I just don't see this claim being able to make advance predictions rather than post hoc cope.
Well, mostly the limits prevent us from even _seeing_ the phenomena that our instruments can't reach. I'll go out on a limb and say that if there is a sharp resonance in particle collisions at, say, 10^16 eV, nothing in our current theories captures it, because our data is limited to about 10^13 eV, and even the most intelligent extrapolation rarely gives correct answers over three orders of magnitude.
> My only point is that that computational/conceptual effort isn't the _only_ thing that is important.
I think this is sufficiently weak that no rationalist, even a "computational supremacist" like Eliezer, would disagree.
I guess my concern is: what is the *use* of this belief, beyond saying the words "experiment matters". I don't think this thinking pattern allows you to come up with more effective experiments, in fact I don't see how it prevents you from e.g. running more experiments when it's time to understand data, because more experiments = more better. If anything the problem in sciences like nutrition or the social sciences is that people are trying to get p values, rather than figuring out what the fundamental facts of the situation are. Like you *say* that we needed a spectroscope, but in my mind the spectroscope was something more like 80% knowledge of quantum mechanics and its implications and 20% "trial and error". If you know specific details of the history of the spectroscope or other instruments which indicates, for example that people were using it for purposes before they acquired understanding of what they were looking at, then I would love to know your thoughts. (This isn't sarcastic, stuff like the Chinese variolating their families with smallpox, most likely the invention of lots of things like smelting were along those patterns, those just feel like much less common and successful than say, establishing a theory of vaccines or of combustion+ air circulation in furnaces.
> Well, mostly the limits prevent us from even _seeing_ the phenomena that our instruments can't reach.
That seems like a strong argument for theory and against naive experimental fatalism. Particle physics is filled with stories like the discovery of the charm quark: https://en.m.wikipedia.org/wiki/Charm_quark
I *feel* as if this type of called shot would be impossible for an experimentalist supremacy world. How is it not?
See "A Study in Scarlet" and the part about the ideal reasoner:
"Then I picked up a magazine from the table and attempted to while away the time with it, while my companion munched silently at his toast. One of the articles had a pencil mark at the heading, and I naturally began to run my eye through it.
Its somewhat ambitious title was “The Book of Life,” and it attempted to show how much an observant man might learn by an accurate and systematic examination of all that came in his way. It struck me as being a remarkable mixture of shrewdness and of absurdity. The reasoning was close and intense, but the deductions appeared to me to be far-fetched and exaggerated. The writer claimed by a momentary expression, a twitch of a muscle or a glance of an eye, to fathom a man’s inmost thoughts. Deceit, according to him, was an impossibility in the case of one trained to observation and analysis. His conclusions were as infallible as so many propositions of Euclid. So startling would his results appear to the uninitiated that until they learned the processes by which he had arrived at them they might well consider him as a necromancer.
“From a drop of water,” said the writer, “a logician could infer the possibility of an Atlantic or a Niagara without having seen or heard of one or the other. So all life is a great chain, the nature of which is known whenever we are shown a single link of it. Like all other arts, the Science of Deduction and Analysis is one which can only be acquired by long and patient study, nor is life long enough to allow any mortal to attain the highest possible perfection in it. Before turning to those moral and mental aspects of the matter which present the greatest difficulties, let the enquirer begin by mastering more elementary problems. Let him, on meeting a fellow-mortal, learn at a glance to distinguish the history of the man, and the trade or profession to which he belongs. Puerile as such an exercise may seem, it sharpens the faculties of observation, and teaches one where to look and what to look for. By a man’s finger nails, by his coat-sleeve, by his boot, by his trouser knees, by the callosities of his forefinger and thumb, by his expression, by his shirt cuffs —by each of these things a man’s calling is plainly revealed. That all united should fail to enlighten the competent enquirer in any case is almost inconceivable.”
“What ineffable twaddle!” I cried, slapping the magazine down on the table, “I never read such rubbish in my life.”
I don't think AI is that ideal reasoner, but I can see the principle of "by sound logic an ideal reasoner could infer the greater from the lesser" and why that's attractive. I don't accept it myself, but I see the position.
Many Thanks! I could see the position, but I also disagree with it. Even from the examples he cites:
>“From a drop of water,” said the writer, “a logician could infer the possibility of an Atlantic or a Niagara without having seen or heard of one or the other.
Good luck understanding the turbulent flow of a Niagara by observing a quiescent drop...
>By a man’s finger nails, by his coat-sleeve, by his boot, by his trouser knees, by the callosities of his forefinger and thumb, by his expression, by his shirt cuffs —by each of these things a man’s calling is plainly revealed.
Good luck telling one office work job from another by such signs...
Yes I loved reading Sherlock Holmes stories as a kid, but of course had to come to a sad conclusion that one can't tell a story of a man's life just looking at his pocket watch.
Of course it was a lot easier back then.
Sherlock: “ I deduce that man dressed in rags and begging on the streets is a beggar, that that sooty child on the roof is a chimney sweep, the bewigged older man with a briefcase mounting the steps of the court house is a lawyer, and the well dressed liveried servant helping people on the coach-and-four is a footman.”
Watson: “The guy in the footman’s attire with the Knightsbridge Omnibus Association livery is, you think, a footman?”
Sherlock.” Precisely. My good fellow. Well noted”
Watson: “literally, no shit Sherlock”.
Sherlock. “Despite the vulgarity of your vocabulary, Watson, I take that to mean that you find me free of excrement rather than “full of it”. I take that as a great compliment”.
Many Thanks, Agreed!
Would you consider yourself one of the people Im criticizing? Believing physics is math? That we solved biology? Modern science should be all statistics?
Many Thanks! No, I think I'm agreeing with you, and disagreeing with them. I'm particularly disagreeing with Yudkowsky's view that with just video of ordinary scenes and enough computation, all of physics could be rediscovered. For biology - ye gods are there a lot of surprises! Consider all of the drug leads, some from computational studies of binding to target proteins, that fail in clinical trials.
In pathogen news:
1. COVID wastewater numbers remain low, but SARS2 is still circulating.
2. COVID deaths and ED visits are down, and the weekly COVID mortality rate is lower than at any point since the beginning of the pandemic (of course, mortality will rise during the next wave). In fact, ED visits are still falling ~11% per week. These generally trail WW numbers by a couple of weeks, so I expect the rate of ED visits will level off in the coming week to reflect the WW curve.
3. In the US, XEC's growth rate has plateaued at ~20% of samples and may be falling now. It's unlikely to cause the next wave in its current form (my opinion). For instance, in the UK
4. Over in the UK XEC is greater than 60%, but COVID cases are falling. XEC seems to outcompete the KP.x variants but doesn't increase case numbers. We've seen this pattern before with some other competitive variants (remember Kraken?). My hypothesis is they outcompete the previous vars in a population that is susceptible to those vars, but they don't infect the rest of the population.
5. OTOH, XEC is taking off in Australia. It's at greater than 50 percent in NSW. And a new wave seems to be revving up Down Under.
6. And I need to retract a claim I made in earlier updates that, during interwave gaps, SARS is circulating at higher levels than Flu and RSV by an order of magnitude (per wastewater numbers). After a little research, I learned that Flu A&B shed significantly less RNA out the gut, and Flu RNA is less stable so even though COVID and Flu could be circulating at similar rates in the US population, the WW would show lower levels of Flu. Calculating actual case load from WW numbers is fools game, but they are good for indicating changes in R.
6. I already posted the latest A(H5) avian influenza news in the last open thread.
7. California reported the first Clade 1 MPox case in US. The patient traveled from Africa. It's unclear if it's 1b—which is (supposedly) more transmissible and more virulent. I'm surprised we haven't seen it here sooner if Clade 1 is transmissible as claimed, though.
8. It's started later than normal this year, but Flu season is starting to take off. Biobot seems to show it's more prevalent in the Western US.
9. RSV rates are starting to rise, as well.
10. Mycoplasma pneumoniae infections are increasing. The CDC doesn't officially track these numbers, but Biofire's proprietary Syndromic Trends site suggest that the current rise in cases may not be as dramatic as the MSM has portrayed it.
11. And finally, an actual SARS-CoV-2 recombination event between two vars in one patient and the subsequent transmission to another was documented. We see a lot of SARS2 recombinant (hybrid) vars in the wild, but this is the first time we've seen it as it happened --> https://tinyurl.com/476bkr2w
My slides are up on ThreadReaderApp...
https://t.co/6WlYGzCGBk
Thanks! And another reason for careful celebration: it's been four years now, and we still haven't seen any cases of the B Yamagata line of influenza. It might be that the worldwide anti-virus measures have accidentally reduced the number of influenza lines from four to three.
Yes, but influenza showed an interesting pattern. Cases dropped off the cliff to near zero *worldwide* just as the pandemic was revving up in March. I wish Substack would allow us to post graphics in our replies because I could show this better than I can describe it, but within three weeks, Flu cases (from all strains) in the US dropped from 18,000 to 1,200, and test positivity rates dropped from ~25% to ~1%. This same pattern was seen in other Northern Hemisphere countries, too. In a normal flu season, there's a long tail out into April, but the 2019-20 Flu season ended early. In China, where COVID was spreading earlier than the rest of the world, Flu season ended in January (with the same precipitous dropoff in test positivity and cases). Although COVID had been circulating in the US since February, the cases didn't really start ramping up until the 3rd week of March. So, the Flu decline in the US preceded the COVID case surge by about 3 weeks (which is longer than the incubation period of SARS2). Several virologists and epidemiologists posited viral interference as the cause of this pattern, but most experts dismissed the idea. The explanation i saw was that the Flu decline was all due to the farsighted NPIs that the US (and other nations) implemented early. But in the US NPIs weren't really introduced until late April when the first wave was peaking (typo correction: I wrote Alpha, but I meant the original A strain). And many countries never implemented NPIs.
Flu was suppressed worldwide in both the northern and southern hemispheres until the end of 2021, when Omicron hit the scene. The northern hemisphere had a late starting 2021-22 Flu season, but overall Flu cases were low compared to previous Flu seasons. The world returned to normal Flu seasons in the southern hemisphere for their 2022 season, and in the northern hemisphere, Flu came back with a vengeance during our 2022-23 season. Note: China had a steady background of Type B cases all through 2021 that peaked early in 2022. This was when their ZeroCOVID policy was in full force and was the most restrictive NPI policy in the world. Then they had a belated burp of Type A in the spring of 2022.
I don't think the NPI and restricted air travel explanations explain these patterns. I suspect there was some sort of viral interference going on. And when it comes to one pathogen affecting another, we can see a weird pattern between SARS2 and Rhinoviruses. In the US they show a noticeable inverse correlation. When COVID is low, Rhinoviruses pick up. And when COVID is high, Rhinoviruses drop. Or maybe it's when Rhinoviruses are high, COVID is low and vice versa.
Finally, B Yamagata cases dropped after the 2017-18 Flu season. And they didn't really pick up again during the 2018-19 season. I suspect B Yamagata didn't have the numbers going into the COVID pandemic to survive the die-off.
Great comment, thank you!
A recent tweet demonstrated a failure on the part of many liberals to grok low-class people:
https://substackcdn.com/image/fetch/f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe33bbcab-239d-4dc2-ba82-7216ed049f5b_1001x532.png
Many have remarked on this seeming paradox. Why does Trump appeal to so many lower-class people despite being a billionaire? It hardly seems paradoxical to the many people in Hollywood whose full-time job is selling an image of luxury to low-class people. The actress partying in her 15 million dollar mansion. The rapper driving his luxury car and bragging about his money. The musician haggling over a multimillion-dollar contract. Of course, the audience needs to be able to relate to the subjects. Such programs focus on the actor, musician, and athlete rather than the derivatives trader because the former are more culturally and psychologically relatable.
Trump has an intuitive understanding of celebrity culture. At the RNC, he didn't invite random people to talk about relatable subjects like not having any money. He invited Hulk Hogan and Kid Rock.
The Democrats' answer was, "Look, we've got a relatable candidate. He's just like you, a middle-aged man who's going bald and wearing a hoodie and has a cat on his lap." And he's "coach." Not the coach of an NFL team; he's the coach of a high school football team. Who trashes J.D. Vance for being a working-class kid who went to Yale instead of Minnesota State University, Mankato:
https://x.com/Tim_Walz/status/1820963380108157166
I think Trump just genuinely likes wrestling. He's attended wrestling events in the past. And I heard he went super into the weeds on MMA (or something?) while on Joe Rogan (I've not watched the podcast episode myself). So Hulk Hogan is not a surprise.
I listened to it, and he did, and he does. Boxing too. (One of the weirdest accusations I continue to hear against Trump is that he's racist, as if to imply the hood-wearing cross-burning sense. Meanwhile, he socialized with black athletes in ways that no hood-wearing cross burner would.)
It's a barber pole thing, or a midwit meme thing.
Low class person: eats McDonalds because it's tasty
PMC striver : doesn't eat McDonald's because that's gross lower class slop
Powerful billionaire: eats McDonalds because it's tasty
The PMC striver class spends their whole time trying to distinguish themselves from the low class yokels, but then demands their loyalty once every four years.
Bad example, MCdonald's is gross lower class slop that costs more calorie for calorie than innout.
Those are good burgers
STFU Donnie
https://m.youtube.com/watch?v=igG4OrblTeQ
I WILL NOT BE SILENCED!
A McDonald's Proleslop patty is good at $.75; at $2.39 you are getting robbed! 300 fucking calories! The bun is at least 150 calories! Fucking HALF!
Never had an In N Out but don’t doubt they are better than McDonalds. You just reminded me of a funny scene in a funny movie
Donnie does say ‘those are good burgers’ after all
Walter says STFU to just about everything Donnie says. Walter lacks the awareness that he is just asshole.
https://m.youtube.com/watch?v=Tsy2CrJaKUs
I've read that some people want to coin a term for the quality of being right, but being an asshole about it, as "Waltersobchakheit".
It's okay, I know, I'm PMC too.
Even during the years my family was homeless and couch surfing, even when I had a 15$ weekly food budget, Mcdonalds was still disgusting overpriced slop.
Innout (or whatever your local innout equivalent is) is right there for delicious fairly priced slop.
...I would start practicing not needlessly painting a target on your back like that.
I will live my truth, and the truth is this:
If you unironically, un-nostalgically enjoy mcdonalds, you are either a child or have lost all your tastebuds in a tragic lit roadflare licking incident.
The ‘tragic road flare licking incident’ line earned a laugh to the point of snorting from Mrs Gunflint. She sends her thanks.
Or you don't live in the US.
> or have lost all your tastebuds
Isn't it the other way around? We lose our capacity to taste as we age - it's why kids are super picky while adults are happier with much more strongly tasting foods than kids are.
We have lost our capacity to appreciate how delicious the McDonalds we loved as kids truly is, because we are old now and those taste receptors are all dead and all that remains is the sensation of eating limp ketchup-covered cardboard.
Eh. People say this, but it doesn't work out that way IMO
You don't loose taste until you are pretty old, but some subset of children are hypersensitive to certain tastes to the point where they destroy their ability to enjoy other tastes.
Eg, I could not eat anything with raw onions until my late teens, to the point where once someone put a raw onion on a hamburger bun, remembered I didn't like them, took it off, and I still picked it out. We actually did a blind test afterwards because it seemed ridiculous and I still got it every time.
Eventually this hypersensitivity faded to the point I could appreciate the astringent, sulfur compound taste/sugars and aromatics and such of onions. I wouldn't say my taste has gotten worse as such, because I can till identify a salad that was tossed in a bowl that had a raw onion in it before, it just doesn't bother me anymore.
My dad explained it to me when I was around ten:
A smart person puts out a sign saying "pedi-habiliments ambidextrously lubricated and illuminated for the infinitesimal remuneration of ten cents per operation".
A *really* smart person puts out a sign saying "shoe shine: ten cents".
Might be a liberal boomer thing. I think they really did grow up in a beatnik culture that rejected looking rich and successful. Times have changed but the same liberal boomers were picking the candidates.
Your first link is correct. Trump had McDonalds served on Air Force One. That's the kind of 'relatable' that all the spin doctors in all the campaigns can't fake.
https://x.com/DonaldJTrumpJr/status/1858136498530374074
And it's not the first time he's done that, and the more stories about being "mocked" for that, the more ordinary people go "yeah, I dunno about him but I sure know *you* are not on my side, so I'm voting for him".
https://www.standard.co.uk/news/world/ivanka-donald-trump-photo-eating-mcdonalds-air-force-one-a4353081.html
As for the second, Walz emits a slight air of jealousy, never mind that his family circumstances at the start were way better than Vance's family circumstances. And I really don't have much of an image of him or what he is really like, because I don't think the Harris campaign did much with him - he was trotted out now and again, but mostly kept in the background not to take any lustre off the Coconut Queen, for all the good it did. Does Walz eat McDonalds? or another fast food chain? Does he do anything at all, apart from "I used to coach high school football"?
I wonder what RFKjr thinks as he decides if he prefers to eat the fastfood or just go hungry. Also the rolled silverware is killing me.
https://theonion.com/sweating-rfk-jr-performs-self-surgery-to-extract-big-mac-from-stomach/
The expression on his face is what sells it. It's clearly not one of the staged "now let's all eat a bacon sandwich to show we're Just Folks", Trump really does like McDonald's and the rest are going along with various levels of enthusiasm.
That's the kind of 'real' that we're talking about in regard to the Harris ads. Did we ever see her eat a Dorito, despite the claim that Doritos are her favourite snack? There was also a photo op with Gretchen Whitmer and the pair of them 'having a beer' together visiting a restaurant in Kalamazoo, but Harris had one sip (granted, you don't want your candidate to get hammered on the campaign trail). At least Whitmer looks like she *would* drink a beer, as she's supping away here:
https://www.youtube.com/watch?v=M-SdPl98mlw
The fact Walz was supposed to embody a new and inspiring model of healthy masculinity is hilarious to me. Who aspires to be Tim Walz?
I will say this for the Harris-Walz... they really do look exactly like Maya Rudolph and Jim Gaffigan. If you wanted to pick a ticket based on the availability of comedians who can play them on SNL, then you could hardly do any better than Harris-Walz.
Trump is terrible for SNL. In fact I've never seen a comedian do a decent impression of Trump, which is incredible.
https://www.youtube.com/shorts/Xj_KXoNnako
How about this one?
That part of it I kinda-sorta get: if you start from the premise that the masculinity we've already got is diseased and has to be replaced with something else, "Dad" is certainly a focal-point solution. The unfathomable part is where the campaign got the notion that they could ever appeal to men by starting from that premise.
He was maximum inoffensiveness, which turned out to be maximum ineffectuality. I suppose the lesson to be learned here, from both Hillary and Kamala's attempts, is "if you want to be First Female Ever, do *not* pick as your running mate a white guy named Tim who is a former state governor" 😁
A good test for Trump's 2nd Term: Try to impound funds for SLS (Space Launch System).
SLS, if you don't know, is an abject joke. Multiple billions of $'s per launch. It's been in development for over a decade. It has launched *once*. With Starship on the field, there's really no reason at all to continue funding this to the tune of billions a year. So: impound the funds.
Now you may say: but!...The Congressional Budget and Impoundment Control Act of 1974 blocks the President from impounding funds! Yeah, you're right. Congress, illegally and unconstitutionally, stripped a power from the executive branch that was recognized as a power of the President going back to the very beginning (to stop Nixon, who had an act to grind with the system of absolute Legislative/Administrative branch supremacy). Therefore, with a mandate to restore Government Efficiency and wrestle back power from the deep state, the President should take this opportunity to impound funds for SLS, and restore the balance of power in FedGov somewhat.
And if the Supreme Court tells him no...he should just do it anyway, of course. I'm sure congress will impeach and remove their own (very popular) president for trying to cancel the porkiest of pork pies congress has mandated we fund. No more!
Impoundment is not an enumerated power of the executive branch. It has a long history of being accepted practice, yes, but that's consistent with it merely being an exercise in discretion within the bounds of the laws and appropriations Congress had established up to that point. The history is also somewhat overstated: the practice in the 19th century was rare and was more often deferring expenditure until a later day or spending less than the authorized amount while still carrying out the program. "Policy" impoundments, where the President uses the power to abolish a program or refuse to enact it in the first place, were mainly a thing between the FDR and Nixon administrations. And there is at least one case (Kendall v. United States, 1838) where SCOTUS struck down executive action around spending (payments to mail contractors) on the grounds that the law regulating the amount to be paid was non-discretionary.
Congress does have enumerated powers to appropriate money for various purposes, as well as to make laws "necessary and proper for carrying into Execution the foregoing Powers". And the President has an enumerated duty to "take Care that the Laws be faithfully executed". Taken together, these seem to authorize Congress to restrict and regulate Presidential discretion over how appropriations are implemented.
>"the practice in the 19th century was rare and was more often deferring expenditure until a later day or spending less than the authorized amount while still carrying out the program"
"spending less than the authorized amount while still carrying out the program" is exactly what I'm proposing he do. I'm not saying to cancel the Artemis program, far from it. Just to NOT do it with the pork project launcher that serves *literally no purpose but pork* (Starship, FH and New Glenn can do everything it can do for 1/10th the price).
You can save $2.8 Billion on Artemis 3 by using Starship and not SLS, and you can reassign the NASA employees to do something actually useful. Artemis 3 and Space Science missions can still be executed faithfully, and everyone can keep their jobs. Make Congress defend this boondoggle of a launcher to the public. If they wanna keep it, impound funds and make the Supreme Court stop you. They probably won't, because despite it not being an enumerated power, using impoundments in a manner like this has historical precedent. If they do, *impound the funds anyway* and make congress impeach and remove you if they really, really, wanna keep SLS. Would they? I doubt it, and that would be fine. This is how the system was designed - it's how it's supposed to work. The President not being to fix garbage like SLS and being totally subservient to congress is *not* how the system was designed, and this is like half the reason the Federal Government is such a bloated mess at the moment.
I'm inclined to agree that SLS is a shitty, overpriced program and should be cancelled. If the appropriation had been written for a generic heavy-lift rocket, then I would agree that the President should have the discretionary authority (and the responsibility) to buy the most cost effective rocket that meets the requirements listed in the appropriation. The constitutionally correct time and manner for the President to object to a wasteful or extravagant appropriation is to veto the appropriations bill.
The problem is, I don't think the appropriation is for a generic heavy-lift rocket. If it's specifically for SLS, and it has been properly passed and signed into law, then the President is legally obligated to buy it unless Congress can be persuaded to amend or repeal the appropriation.
As a more general question, you seem to be arguing that the Presidency is much, much weaker relative to Congress than the Farmers had intended. This is a rather novel interpretation of affairs -- I have much more frequently heard the reverse argued.
The President isn't allowed to make decisions on where to spend money, with a few exceptions for emergencies and such. That is a power of Congress: setting budgets and allocating money. The President is the chief executive, to carry out laws. If the President can decide to spend money, or not to spend money, then Congress no longer has that power. This is why Biden wasn't allowed to forgive student loans.
If Congress wants to spend billions of dollars launching things into space, the President must ensure it gets executed. The particulars as to WHERE the money is spent is, indeed, up to the President. But if funding is to be pulled, Congress must do it.
Impoundment is just NOT spending the money congress gave him to spend. Very different from authorizing spending where there was none previously (like the Biden student loan thing).
Congress is allowed the power of the purse, that much is true, but the president deciding "nah" was within his power til Nixon, especially in cases where the President is still executing the law, just more cheaply than expected. Doing Artemis with Starship only fits the bill.
The SLS should be cancelled, even if Starship isn't ready to replace it. But the problem for Trump is the SLS was supported by Republican senators and maintains jobs in Republican states. So he'll need to go against his own party to cancel it.
From what I understand, it was mainly Richard Shelby, who's gone now. Most of the jobs impact (were he to cut the program completely, rather than have them reassigned to something else) would be in Alabama and California. Alabama is red enough he can tell them to pound sand, and the rest of the party is not gonna go against Trump + Elon to save a useless pork project.
The problem is that with Elon in charge of pork-cutting, it will look very bad to kill of a competitor to SpaceX, even if objectively valid.
Calling SLS a competitor to SpaceX is absurd on its face. It literally only exists because NASA mandated it to a decade ago. It serves no role that Starship or New Glenn couldn't do better / cheaper.
I don't think he gives a damn about what the left thinks of him at this point. The right that's loyal to Trump will probably be fine with it, for privitization/cost-cutting/owning-the-libs reasons.
Trump needs to pick his battles, he doesn't have infinite capital to burn, especially when it comes to pissing off red state senators.
He has something to offer those red state senators that no other Republican or Democrat can offer: power. Real power. That's all the leverage he needs.
Sure, but power isn't a limited resource even for a President, and every bit of power you trade for one thing is a bit of power you can't trade for something else.
> It's eerily similar to the good times, weak men meme.
But in reverse right? Good times come from generous, cooperative bots (i.e. 'weak'), but eventually everyone is so generous that bad bots take advantage of them, leading to a period of bad times.
I think the cycle he was referring to is
1. Bad times (with lots of defecting bots) create strong bots (tit for tat takes over).
2. Strong bots (tit for tat) create good times.
3. Good times create weak bots (cooperators take over from tit-for-tat).
4. Weak bots create bad times (defecting bots take over).
The "Win-Stay, Lose-Shift" strategy breaks the cycle by punishing the cooperators that allow themselves to be eaten.
https://tsvibt.blogspot.com/2024/09/the-moral-obligation-not-to-be-eaten.html
> 3. Good times create weak bots (cooperators take over from tit-for-tat).
Under what thoery and game variant does this ever happen?
When there's a randomized component of the game, where bots might "accidentally" play the opposite of what they intend. TfT can get into an eternal grudge match while always-cooperate recovers smoothly.
See https://ncase.me/trust/
tit for N tat is still better and im pretty sure you can calculate the best possible N from an unknown failure rate
I'm not sure if its tested by actual game theorists, but I could imagine one where you take a tiny penalty for each line of code. So tit for tat looses to pure cooperate if those are the only two players.
Penalizing "thinking" (slightly) has crossed my mind as well as a worthy addition to a simulator. TFT has that nice property where you only have to keep one register for the other player. And this feels intuitively like what a lot of people do. (Trump is famously called out here - he seems to only remember the last thing you said about him, and if it was positive, it doesn't matter how much you criticized him before that.)
So, we could incur a small penalty for memory (do I remember what the last move was against me? Do I track each player's last move? Do I track *all* the moves? And the payouts?), and for processing (how much analysis do I do about each move and payout?).
A simulator might have players little smarter than an amoeba, where TFT or WSLS dominate, or smarter players where the dominant strategies start to resemble stock trading algorithms. Or a mix, where the dominant player might implement nearly any strategy at all and dominates anyway because it plays so many games.
tit for tat is a single bit of storage and single operation; tit for N tats is also tiny and can be selected for many rule variants. If a single bit of computation is to expensive, always defect is also computeless and kills always cooperate.
For 1v1 matchups if a single bit is expensive. Always defect beats always cooperate. Tit for Tat beats always defect (the value of the strategy > the expense of computation) and always cooperate beats tit-for-tat.
If we think of humans rather than bots, the time and effort it takes to learn the skills of whose a defector and whose sob story is completely true means that in a world where most people are honest you'll get a lot of cooperate bots.
> For 1v1 matchups... Tit for Tat beats always defect
No, tit for tat needs an ecosystem and a friend to evolve in
> the time and effort it takes to learn the skills
After a single potential defection, a learning strategy should invest, no?
Where is the appetite for Milei style economic reforms in the US coming from? Both Musk and Ramaswarmy have said America needs such a shock reform followed by a recession and economic pain, but why? Argentina's problems are distinct and not at all like America's. Shock therapy seems to cause unnecessary pain when the problem simply isn't that acute and there are less painful alternatives available. Or do I miss some deep underlying problem that can only be cured by a shock?
It’s class warfare with the signs reversed, that’s all. These policies are hideously unpopular with regular voters of both parties, who want the government to keep its hands off their Medicare/Social Security/farm subsidies/etc.
Cruelty is the point. In Argentina's case, some suffering in likely unavoidable, just like in the Eastern European countries coming out of the Soviet-style socialism. The US doesn't need the same medicine Argentina needs, but both Trump and Musk seem to like hurting people, and Ramaswamy... nevermind, better not flirting with bans, so there.
It's just basic LET'S DO THINGS! style populism. What is being done is unimportant, the important is that it's hard-charging, and drastic, and radical, and that it will fix everything! Wih another set of memes they'd be advocating for the state to DO THINGS! directly.
Musk (re)posting various studies with headlines that sound briefly silly until one thinks more than a few seconds (eg. like https://pbs.twimg.com/media/GcbbzANWkAEofik?format=jpg&name=medium and https://pbs.twimg.com/media/GcbcSDiXUAA83uu?format=jpg&name=medium) rather moves my priors to the direction that what gets done is less Milei-style reforms and more like Golden Fleece style nibbling at the edges. I don't think that drastic Milei-style cuts would be good, but neither would be just turning this exercise into potentially useful research funding being shot down because lol sexbeetles.
The US has some longstanding and well known financial issues that could have been fixed previously but every year gets worse instead of even staying neutral. The national debt increases regularly and the amount of increases keep growing instead of shrinking. We're on track to be paying a trillion dollars a year on interest to the debt next year. Social Security is on an obvious track to insolvency.
There are pretty good reasons why we haven't fixed these issues, but they obviously need to be fixed. It's unsustainable, and everyone knows it. We're unfortunately playing hot potato with these things now, because whichever party actually fixes it is risking not getting elected for a long time, since it is likely to be very unpopular. Milei ran on a platform of doing things to fix long standing problems even though it will cause a lot of pain. If Trump's team can pull that off, then we might dodge the twin bullets of financial ruin and political ruin.
Trump actually increased debt in his first term, and I don't think his tax cutting policies will change this time around.
No disagreement here. The question was about the appetite for Milei-style cuts, which Trump is signaling he would be willing to do. I also agree he will find it difficult to pull this off, and will be reluctant to actually try. It will depend significantly on whether Trump is willing to give real power to someone near him (while Milei is the ideologue who believes in what he's doing).
Past practice shows that Trump is unlikely to do that, but I'm trying to be optimistic that things may be changing enough to get it done.
Sure, but cutting taxes is sort of the opposite of reducing the national debt.
The last time the federal government ran a surplus was under Clinton, who worked out a compromise with the GOP to raise taxes and cut spending. Bush inheritied an 100b annual surplus which he blew up with massive tax cuts. Obama worked out a Clinton-like compromise with GOP Speaker Bohner but the House GOP rank and file were dead set against giving Obama anything, so it died before even coming up for a vote. Now we are going to have even bigger deficits. I don't see any possibility of a fix without both tax increases and spending cuts.
>Clinton, who worked out a compromise with the GOP to raise taxes and cut spending.
Clinton and Congressional Democrats raised taxes a lot and cut spending a little in 1993, with zero Republican votes in Congress. Then Congressional Republicans pushed large spending cuts and small tax cuts in 1995-1996 against Clinton's objections. Clinton and Congressional Republicans did compromise on something that still cut both spending and taxes, but by less than Republicans had wanted. That compromise was projected to balance the budget by 2002, but windfall tax revenues from the Dot Com boom lead to the surplus appearing early, in the 1998 fiscal year.
The combined effect was a combination of tax increases and spending cuts, as the tax cuts (even before Clinton pushed back and got them reduced) were a lot smaller than the 1993 tax increases, but it's important to note came in two stages with different Congresses approving each stage.
>Bush inheritied an 100b annual surplus which he blew up with massive tax cuts.
Also massive spending increases, both domestic spending (mostly Medicare and Education) and war. And the Dot Com boom ended. We could have afforded any two of the tax cuts, the domestic spending increases, and the wars and still had a surplus again by the time the recession ended, but not all three at the same time.
IIRC, Clinton also cut the deficit by borrowing from the Social Security fund in a manner that didn't come due until the Bush administration. (Imagine if Gore had won and had this dropped in his lap.)
You don’t recall correctly. You are probably remembering somebody arguing that the unified budget numbers were misleading. The argument is reasonable, which is why the Social Security program is “off budget” in the first place, but is as applicable to the Bush years as the Clinton years. There was nothing to “drop into the lap” of a hypothetical Gore Administration.
The “on budget” numbers exclude “off budget” programs, including Social Security. The “on budget” numbers don’t distinguish between debt held by the Social Security Trust Fund an debt held by the public. For that reason, it is not possible to cut the “on budget” deficit by borrowing from Social Security
The “unified budget” includes all Federal programs, including “off budget” ones like Social Security. During the Reagan Administration, Congress passed a plan prosed by the Greenspan Commission that caused the Social Security Trust Fund to run a surplus in order to build up funds to pay for the baby boom retirees. This resulted in the unified budget showing a much smaller deficit or larger surplus than the on budget numbers. In fiscal year 2000, for example, the unified budget showed a surplus of $236 billion whereas the on budget numbers showed a surplus of only $86 billion, mostly due to Social Security running a surplus. This discrepancy increased during the Bush Administration; in fiscal 2008 the on budget deficit was $642 billion but the unified budget had a deficit of only $459 billion. So the unified budget deficit was reduced due to borrowing from Social Security, but Clinton had nothing to do with that.
I don't quite understand what you mean by "on budget". Looking it up, it appears to mean what I took to be the plain reading, which is spending within prior limits, as opposed to unforeseen spending. If so, SS ought to be on budget, not off, and you say it is. So that's probably not what you mean. The meaning I found seems closer to what you're calling "unified budget".
If I just move on to your last paragraph (where most of the meat is anyway), I have trouble tracking down the legislation you cite. I see the "Social Security Amendments of 1983", aka H.R.1900 (https://www.govtrack.us/congress/bills/98/hr1900/text), but my skim of it doesn't tell me anything obvious. ("Surplus" doesn't appear anywhere in the text, so if it happens, it's implied, either by raising FICA or lowering payouts, and indeed, a lot of the bill does that, so, maybe?) I can see how your account explains an apparent reduction of deficit / increase in surplus. It's possible the explanation I heard was referring to HR1900 (or whatever the proposal became), but I can't tell for sure.
If I happen to run into someone who argues that Clinton merely borrowed from SS, though, I can pass this along and see if it jibes with their evidence.
I do notice, meanwhile, that if this plan happened for the reasons you describe, then it wasn't general fiscal prudence on Clinton's part, but rather a specific concern for the Boomer retirement surge that went back to Reagan, which is new to me.
>I don't see any possibility of a fix without both tax increases and spending cuts.
And the former is not going to happen. It is unlikely that the Republicans will even permit the 2017 tax cuts to expire as scheduled at the end of 2025.
I have ended up following more US politics than local politics without even trying. Does anyone have a good writer to follow for succint, rational takes on the state of European/German politics?
https://www.eurointelligence.com/ might suit you.
This does seem promising. Thanks.
What do you want to know?
Very simplified but: Germany is having economic issues. The German government wanted to spend more (mostly domestically) and did its equivalent of hitting a debt ceiling (in this case a debt brake law). Scholz fired his finance minister after the finance minister's party didn't agree to either increasing taxes or changing the debt brake. This led to his coalition no longer having a majority in the legislature. So he called an election to sort it out.
Overall it looks like the Center Right is going to win right now. Though they will need to form a coalition and with whom is an interesting question. But we'll have to wait for the vote totals which we should get in February.
I want to know where I can go for reasonable succinct takes when there's something going on here, not just now.
I'd also like that!
Say there's a magic button that, when you press it, reduces your LDL cholesterol by 30%.
It stays like that, no evil genie stuff.
How much would you pay to press the button?
> Say there's a magic button that, when you press it, reduces your LDL cholesterol by 30%.
I really hope this is market research, because having a probiotic you could take that would reduce LDL by 30% would be awesome.
If I were trying to sell it, I wouldn't even want to get it FDA approved, I'd just mass produce Cholester-yogurt or whatever and hedge with a bunch of "this MAY reduce cholesterol, as part of a complete breakfast" or whatever, like the oatmeal people do, and let people do their own research and make their own conclusions.
But then, I've always been a fan of "go high volume and small margin" as a strategy versus the usual drug / bio strategy of "niche and high end."
You've hit the nail on the head, my guy! The good lawyers at General Mills have spent decades carving out the Cheerios Exemption, AKA Structure/Function Claims.
Probiotics are foods/supplements in the US, so the primary regulatory burden for a new food is just proving safety as long as you don't make disease claims. Cholesterol is really in the sweet spot here because everyone knows that "number go down" means lower risk of heart attack and stroke.
And yes, I'm hoping to make this as widely available as possible, but the logistics involved in anaerobes-as-probiotics mean substantial running costs and geographically limited rollout
Ooh, exciting!
I hope you'll publicize it on your Substack whenever it's ready. I'd probably pay $1k, and I've got no history of triglyceride problems or CVD in the family.
No idea. Not nothing because I have a. Just been reading a book by a woman who made herself tetraplegic falling off a horse and b. Learned that spinal column strokes can have the same effect. But the question isn't answerable unless you quantify the changes in absolute and relative risk caused by pushing your button. In the most favourable case perhaps £10000? I am oldish and have high LDL for which I get free statins from socialised medicine.
My copay after insurance for the lowest possible dose of rosuvastatin is $6 a month. My LDL wasn’t really that high but the statin reduced it more than 30%.
Sorry to anyone who isn’t this lucky.
$4855
lol how'd you get there? "just vibes" is an acceptable answer but I'm curious if there was math.
I had a physical a couple months ago and doc said my cholesterol was high. By about 25%. Since then I’ve been trying to eat right. I eat walnuts and fucking mackerel all day. I’m losing it.
Statins are a no go because I’m religious about only taking drugs that get me high.
If someone said they could perform surgery to do what your button did, I’d probably get it done if my plan covered it. If all I had to do was press your button, no side effects, and witness medical magic, $4855 seems right, but not a penny more.
***
Walnuts and mackerel *every* day? I like mackerel but even that is a bit much 😁
God alone knows what are cholesterol reducing foods, I've eaten nearly everything suggested and nothing worked (niacin helped a bit, but that was the "your face starts burning like a cherub after taking it" version and I couldn't tolerate that long-term).
I’m guessing it was a wise guy answer to a demand for a specific number on a pretty contrived hypothetical.
Somewhere between $500 and $1,000.
As for the question, how much I would pay would depend on several factors
• baseline LDL levels
• risk of heart disease, stroke, diabetes etc. (at the time of determining how much I’m willing to pay)
• disposable income
Right, I'm saying you, personally, right now. Given your risk of heart disease and your disposable income.
And yes, statins exist, but plenty of people on statins still have heart attacks or strokes. I suppose, if you take statins now and wouldn't have to after pressing the button, you could calculate a floor value based on how much you'd expect to pay for statins over the course of your lifetime—but risk of cardiac events is basically a function of the integral of LDL with respect to time, so even someone who's still gonna have to take statins should theoretically derive a few QALY of benefit from the button.
Doesn’t something like this already exist? ie. statins
Sure but statins are a thing you have to keep taking.
One of the things I dread about getting older is the idea of becoming a Person Who Takes Medication. Once you enter that category there's no getting out of it, it's just an ever increasing daily pill load until you're dead. I'd definitely pay the $4855 to put off needing to enter that category.
This is an oddly specific medical conditionality, both the condition and the amount.
It seems to me the benefit to this would allow you to eat anything you want without the problems associated with higher LDL cholesterol. As such, it equates to how much you're missing from not eating those things. In my case, it might be worth something like $100 a year.
I don't really think I ought to disclose my medical conditions and medications in this format, so it's tough for you to see my context.
My LDL is fine. My total cholesterol is fine. But my triglycerides are way too high and I can't shift them. So I'd pay a reasonable amount for that kind of reduction.
What counts as "reasonable" is up for debate.
I want a number! No wrong answer, but half the point of this question is to figure out how accurately people value their lives, given that heart disease is still the #1 cause of death.
I value my life very highly but don't have much of an idea how much lowering my LDL is going to extend it, in isolation.
I could say it would be worth €1 million, or €10 million, or €100 million to me. I don't have anything approaching that kind of money, so it's a useless question.
If you mean "out of the money you have right now, as a once-off payment, how much?", then I'll pay €100 as that is literally the maximum I can afford. EDIT: This does not mean I only value my life at €100, I mean that's all the spare money I can give for something not immediately urgent like the house burning down. See what I mean about stupid questions?
Now, if you're finished asking dumb questions and getting dumb answers, why not ask people "would you take this? would you value the opportunity?" instead of this mania for "put a number on it! it don't mean nothin' if you don't put a number on it!" for the rationalists here. If you want numbers, heck, let me throw the US interest of $1 trillion dollars (as quoted in a different comment) on it as the kind of meaningless dumb valuation of "put a number on it" - yes, I'd very much like to get my triglyceride levels down since the actions and medication I've been taking to date haven't helped. If a magic button fixed that, I would be happy to press the magic button.
>This does not mean I only value my life at €100, I mean that's all the spare money I can give for something not immediately urgent like the house burning down.
No, see, this is helpful. Like, everyone values their life more than their possessions, but that's the thing about uncertainty and probability—you never know that you should've prioritized "not having a stroke" higher until you're in rehab trying to re-learn how to talk. I expect if you were somebody who'd already had his first heart attack, you might be willing, for instance, to sell your phone, or take out a small loan so you could pay €200 or €300 to hit the button.
FYI, I'm not just doing thought experiments here, I'm doing market research. Genuinely trying to figure out how to price a one-time intervention with potential to persistently lower cholesterol. No idea if it works for triglycerides yet, but if you want I'll keep you posted.