ndriscoll

(At least some) error-correcting codes are based on polynomials over finite fields. I couldn't say much more, but it's at least intuitively plausible since e.g. an nth degree polynomial is defined by any n+1 points, so if you know say n+1+p ("p" for "parity") points, you can lose up to p and still recover the polynomial.

Nope, math notations are optimized for reading, not writing (consider that people still use symbols on computers despite it being quite a bit more tedious to type). The conciseness makes it easier for you to see structural patterns and do symbolic manipulation in your mind's eye. Even something basic like the wave equation would become completely illegible with an expanded notation like that.

Same reason why we write 5-3, not subtract(minuend=five, subtrahend=three).

Civilization (at least 4) also has subjugation as an explicit mechanic where a mostly defeated AI state surrenders as your vassal and has to give you resources, vote for you in UN elections, fight in your wars, etc. but they otherwise continue to act independently.

The point is if you're renting, you're already paying for those things while also introducing a middleman and losing autonomy. There's no reason why renting "should" be a better choice even on a relatively short timeframe. e.g. if more were automated to reduce transaction costs, and if the housing market were more stable, then it seems like it'd obviously be better for society for people to own even if only for a few years. On longer timeframes, we should certainly want people to own; we want people to view their surroundings as something they should take care of.

I didn't vote when I rented in San Francisco for a few years because I had no intention to stay, and thought it immoral to disenfranchise the locals. Actually in my opinion you shouldn't be able to vote until you've lived somewhere for at least 5-10 years. Certainly even if you can, you shouldn't if you know you're only there temporarily.

I don't see why you think owning should only be slightly better after 10 years; if you don't want to deal with maintenance directly, you could always hire a management company as if you were the landlord and renter, but there's a benefit to you and everyone around you when you are invested in your home and community.

It's probably more that hardware and software were still rapidly getting better in 2004, and HL2 looked significantly better than HL1. By contrast, just looking up the two on youtube, Doom the Dark Ages on ps5 pro isn't obviously any better looking than, say, Skyrim (2011). Even Morrowind looks pretty good with some high resolution pbr textures and a few new shaders.

The commands don't require you to know the underlying storage model. e.g. I have maybe once in 15 years had an occasion to think about packfiles. I couldn't tell you basic facts about the storage like whether a git clone will mean you necessarily have the same object files as the upstream (I presume no). Content addressed (with named references) snapshots aren't the underlying storage model. They are the UI model, and they make a ton of sense for that purpose.

Merge conflicts don't put you into a detached head state. You stay on the branch where you ran git merge. Why would it change your checked out branch?

The special command is just `git add` to stage your changes followed by `git commit` (or do it all in one command as `git commit - a`), same as any other changes.

Rebase is something else. It's something you do on purpose because you specifically want to rewrite the history. Personally I use it a lot and expect everyone on my team to, but you can use git just fine without ever rebasing or even knowing it's there.

Like I actually have no idea what you're talking about. Is your criticism actually targeted at some specific UI tool that's not git?

It is inherently simple? It's a couple tools to manipulate a graph of snapshots. You can make commits, checkout commits, make/move references to commits, diff commits, and apply diffs to commits. If you wrap your head around that you've pretty much mastered it.

It doesn't "drop you into an interactive rebase" unless you tell it to. If you want to rebase, rebase. If you want to merge, merge. As you say, even most UIs for git show what you are saying you want.

https://youtube.com/watch?v=PlfOoKTFcAE seems pretty good to me?

This all has nothing to do with the fact that the comment I replied to is suggesting that somehow actually being better is not meritorious. If your parents, grandparents, and community did know how to raise you better, and you turn out better, you're still better.

The point is we should not point to that and say "you didn't really earn that;" we should point and say "here is the example for others to look to. Everyone be more like them."

Like I had a broken home growing up. The valedictorian had an intact family in a higher social class. I put in no effort, was rude to people for no reason, and was generally not a great person. He worked hard, was friendly to everyone, and was generally a better person along pretty much any metric.

The takeaway from that is not to think "wow that wasn't fair," but to think "how can I try to be more like his parents for my kids?"

In what way is that not meritocratic? The goal isn't to find the most Christlike figure born to the humblest beginnings and tempted thoughout his life by literally Satan to fall off the good path; it's to find the actually realized best qualified person.

If you're the smartest, most knowledgeable, most conscientious, most humble person because you had a fantastic support system, well you're still genuinely a great person. If you are a D student with a felony criminal record because everyone around you was a horrible influence and your parents abused you, you're still genuinely terrible.

I've seen parents in a store together where one just "supervises" their toddler watching youtube or something while the other shops. Meanwhile I'm there reading to my kids while my wife shops. I'll not be surprised if my kids have better outcomes in large part due to the family they were born into, because I do that on purpose. Society should incentivize people to devote ourselves to raising the best next generation we can. Trying to define merit via some counterfactual hypothetical person instead of the actual person in front of you makes no sense. In fact it's counterproductive as you end up incentivizing people to do things like send their kids to worse schools so they can appear relatively better.

Alternatively, products would be cheaper as we'd legislate away defection in that particular instance of the prisoners dilemma.

Worries about particulars like infinite scroll miss the point: fundamentally, the advertising business model is a market distorting one that companies use to provide goods and services below cost. It is paid advertising that should be banned. Doing so would eliminate the incentive to create addictive products in the first place since increased usage would no longer drive increased revenue.

OP is about information, not functionality. In the early 2000s you would put things like that on a web page, and you'd put e.g. chat in its own application like Gaim.

In the 2010s the model inverted: now you need to keep an entire browser open to use google chat, and people try to get you to install an app to read a web page.

Not that I think this significantly alters the point, but it's pretty common in the US to regulate or ban signage. e.g. billboards are illegal in my city and there are specific regulations about what kind of elements can be present on buildings to signal business names. I'm pretty certain illuminated signs beaming into people's homes would be illegal here. Actually I don't think light-up signs are allowed at all; I believe they have to be lit via projected light pointed at the building the're on.

Couldn't you just instruct the model to always use your tool call to spawn subagents? Subagents are not some magical thing; it's just another prompt with a couple tool calls for plumbing. One of my colleagues made his own subagent harness earlier this year before codex had them at all.

I didn't say humanities majors don't earn more. People who earn more with their degree aren't the contention. The contention is the nebulous, intrinsic value of a "more educated society", c.f. https://news.ycombinator.com/item?id=48880930

The constitution might've been mandated as part of the curriculum. That doesn't say much. The question in this subthread was about the relation between "education received" and "amount of knowledge possessed", to which I replied that I doubt most American college educated adults have actually read the constitution.

To illustrate, I had never actually bothered to read the constitution at the time that I got a 4 on the AP US History Exam. It is not hard to pass tests without reading or thinking about the material, and when it comes to what's part of state standards, we all know it's nearly impossible to fail US K-12.

So I am talking about all of those things; the constitution and Locke and Shakespeare and Homer are all supposed to be on the curriculum. How many people actually read them? To what extent is even more education a delusion? If you don't need expensive lab equipment (and this includes some STEM fields like math/CS), do we need everyone going to universities? Or if you actually want to learn the material, can you just do so?

The original post defined it:

The test itself is simple: If an undergraduate program's graduates don't earn more than workers who never went to college, that program could be cut off from federal student loans. The same goes for any graduate program whose graduates earn less than someone with only a bachelor's degree.

I'm not going to go so far as to claim that there's a widespread project to undermine our society (as opposed to individual indifference or disdain), but you yourself demonstrated the issue: when I express my doubts that supposedly educated Americans have ever actually read their own constitution, you ask why that's an issue an why we should privilege that above other reading. Ancient Chinese political theory is neat and all but how about knowing the basics of our own country?

The post explicitly calls out that solution manuals are available and that one should do most/all problems. For highly mathematical topics you can also do derivations (or proofs in math) yourself before reading the one in the book.

People earning higher wages as the result of their degree are explicitly and exactly not the topic of discussion. The discussion is about funding for programs that measurably do not produce economic benefits, and some supposed intangible benefit to just being educated.

I get the conclusion from having been a student, having socialized with other students, observing discussions from teachers online and seeing that things are either the same or worse, and observing my fellow citizens. It's like asking how I know most young people in the US have premarital sex or how I know the obesity epidemic isn't actually a bodybuilder epidemic: because I live here. If studies said otherwise, I'd have to assume the study was wrong.

STEM education is easy to demonstrate the utility of because it shows how to make things that improve our lives. People working at Instagram should face criminal stalking charges, and gambling (or at least advertising it) should likely be made widely illegal again. That doesn't mean people programming inventory management systems or ensuring bridges have adequate safety factors and avoid catastrophic resonance aren't directly providing value with skills they learned.

I suppose it depends on what you mean by unstructured (vs. the original claim of self-study). e.g. my method of self-study was generally to find course requirements for a degree program, find corresponding course descriptions and syllabi, find the relevant books, course notes, and/or video lectures and proceed through those.

e.g. I did formal study of basic physics in an engineering program, but Noether's theorem was never mentioned. I came to that through self-study, and can't really imagine how anyone would miss it unless "self-study" means "read random blogs and watch random youtube videos." I expect even browsing Stack Exchange and Wikipedia would expose you to most core topics fairly quickly.

I'm not making nearly that strong of an argument, and indeed neither is the regulation under discussion. They need to demonstrate that it's somehow beneficial on average to economic outcomes. That could mean:

  * You were able to apply that knowledge to something someone wanted. 
  * The experience somehow inspired you to find your own path to success. 
  * Your degree helped convince someone else that you're a generally intelligent or knowledgeable person and therefore more valuable to them. 

I take your friends' work to mean they're in some academic position? Because funding some small number of those is a different proposition from funding a significant portion of the population to study whatever.

Though actually if we back up a bit here, my original point was more that I doubt people are even learning things. e.g. if your history friend is unfamiliar with important parts of their own country's history (stuff that might even be expected in high school or earlier, but people don't do the readings), I'd be skeptical of the value of their program in the same way that I'd be skeptical of an engineer that doesn't know the difference between an ionic and covalent bond, or Newton's laws, or some basic programming, or RLC circuits, or other basic things that someone technical should know.

Self-governance and self-determination are the founding assumptions underlying our democracy. This is pretty clearly stated in our founding documents. This has nothing to do with capitalism.

I agree that people at Instagram and DraftKings are negatives to society, and probably they also failed to obtain their liberal arts education. Largely I think people with these things on their resume should be blacklisted by others as they proudly wear their lack of morals, and we should do more to regulate such things.

Maybe my wording was off; by textbook my meaning was more along the lines of treatise, i.e. not casual.

You can place them in a dialog with other texts, but that implies that students are reading and becoming intimately familiar with the canon being critiqued. And generally speaking, I'd expect more of a refinement than a critique per se.

I didn't say we should forcefully propagate anything; I said we shouldn't fund things that aren't useful somehow. Some basic values would be that e.g. governments derive their authority from the citizenry, and should therefore always act in the interests of the citizens. If the citizens aren't seeing some benefit, then they shouldn't be taxed to fund a thing. Assuming people disagree on whether they want something, it follows that government should be as local as possible, so if we do fund such things, we shouldn't do so at the federal level unless there is widespread agreement (e.g. maybe we all agree to fund STEM degrees for their economic utility but there is disagreement on social science so those are done locally).

Or that as a high level idea, our entire system is predicated on the idea that people are qualified and able to define their own destiny and self-govern, which is generally in opposition to the idea that we are determined by various privileges and circumstances that require all sorts of compensatory adjustments. Basically, if you want to flourish, just go do that. Again, the books are all there. Most if not all of them are free. We already have 13 years of free, compulsory schooling to learn how to read them.

Or the idea that rights should be principled (e.g. carry some symmetry), not arbitrary. c.f. elsewhere in the thread about natural rights. And by extension, that some systems (e.g. ours) are better than others.

The context here is targeting programs that do not produce economic benefit, so that is hobby study, yes. There's nothing wrong with hobby study. I read textbooks all the time for fun. I think we should encourage everyone to do a lot more of it instead of e.g. Instagram. But I also have a day job.

I'm mostly speaking to the students, but certainly there are faculty that would apply to as well.

If we're talking about some intangible benefit to society for having lots of educated people, I would interpret that as being in service to civic contributions, so yes I would expect that means a liberal arts education should teach Western canon, and certainly it should not undermine it. If not in service to economic utility or propagating our values and traditions, then why exactly would we support a large portion of the population to do nothing but pursue hobby study in something no one else wants for years? Actually it's not even clear that the students even want to do that study either, given the apparent mass cheating or AI usage. So who is it for?

Right but that's just it: it's barely the upper half (over 40% of young people now get a degree apparently[0]), and the age breakdown indicates that 40% is also apparently the proportion of 18-29s that know we declared independence 250 years ago. So assuming continuity and rough correspondence of the two distributions, that's approximately the level of knowledge at the threshold of degree holders now.

Actually the census data says ~43% get a degree and the article says 39% of 18-29s know what we celebrated, so presumably at least 4% (~9% of degree holders) are able to get a degree and don't know.

If this is our bottom end, I get the sense that even above-average students (even of liberal arts) aren't reading Montesquieu or whatever image people try to evoke about highly educated societies.

[0] https://www.census.gov/newsroom/press-releases/2025/educatio...

They might be part of the curriculum, but apparently ~half the country doesn't know what we just celebrated[0], so I somehow doubt they actually did their reading. It's pretty easy to pass or even do well in classes without really engaging with the material at all, even at the university level. Like I got an A in a philosophy course by just reading the professor's notes on exam morning and spitting them back out on an essay. I've still never gone back and actually read Grotius.

[0] https://www.npr.org/2026/07/03/nx-s1-5881451/cato-institute-...

I don't think that's really a counterpoint. My point was that I get the impression that even those in the humanities seem to be losing familiarity with liberal ideas. So it's even worse now (or we've just lied to ourselves as we pretend to expand access to higher education).

To that point, how many university educated American adults have actually read, say, Locke? It's been freely and easily accessible for decades, e.g.:

https://www.gutenberg.org/ebooks/7370

How many university (liberal arts even) educated Americans have read the Federalist Papers? The Declaration of Independence? The Constitution?