bzbarsky

I think it's important to believe you can improve, and that most people can improve with practice. This avoids the very real problem of people self-limiting in unnecessary ways.

I think that telling people people that they can achieve _arbitrary_ things through enough effort is just wrong: it sets up unrealistic expectations and makes them blame themselves when things don't work out, because they must not have tried enough. Telling someone who's 5 feet tall as an adult and not particularly athletic that if they just work hard enough they can become a successful Olympic swimmer is not OK.

It's a tough communication problem, because "maybe you can reach this goal if you work really hard and happen to be lucky too" is a much harder concept to communicate and internalize than "you can reach this goal if you just work hard". It's also a tough problem in trying to figure out what expectations ought to be "realistic" for a given child, because of all the biases we bring to such evaluations. So erring somewhat on the side of emphasizing growth mindset can make sense, and should absolutely be done when doing evaluation, but you have to be pretty careful how you communicate the resulting recommendations.

Past that, the unfortunate thing is that Dweck's disciples (and I use the word specifically for its religious connotations) go overboard on the "everything is possible with a growth mindset" thing, in exactly the way that I think is harmful. I see this in a lot of elementary school teachers; they love this book, but end up in the "growth mindset solves everything" trap, at least in their communication.

I should note that unlike many things in that field Dweck's work actually replicates in large pre-registered studies, albeit ones still done by Dweck herself. The effects are quite small on average, but maybe on the margin (i.e. for students who are more likely to doubt themselves for various reasons) might be significant.

That said, I would be happier if someone who is not Dweck were able to reproduce her results, and I have ~0 confidence that "growth mindset as practiced in elementary schools today" is at all useful...

I think you're spot on, for what it's worth. The context for my comment above is that I have a math PhD, and even then a lot of the articles are an unnecessarily tough read. For the general public they are pretty impenetrable...

If I had examples off the top of my head, I would have cited them. I'll see if I can find time to do some digging through my browser history, but this is a multi-hour endeavor that you're asking for here, so I might not be able to get to it.

And to be clear, I am not talking about people staking out territory or looking for recognition. I am talking about articles being so badly written they are functionally indistinguishable from being "wrong". Yes, you can decipher them with enough outside knowledge and some other references, but it's pretty tough. And this is not a majority of mathematics articles by any means. But the problem is that if you don't know enough you can't tell whether you're looking at such an article or not, unfortunately...

P.S. I just took a quick look at just my recent-ish Wikipedia edits, and https://en.wikipedia.org/w/index.php?title=Limit_superior_an... was correcting a small but obviously wrong claim. Which is, to be clear, not the sort of issue I was referring to in my comment about badly-written articles.

No, when I say "limited government" I mean just that. A government that has certain powers, but not others.

This is not just about enforcement of laws, it's about which laws are allowed to exist in the first place, and about government procedures adhering to those laws.

To give you concrete examples:

* I don't think "the government" (broadly defined) should have the power to stick some poor kid in Riker's Island for a few years while they think about getting around to trying them on a trumped-up charge.

* I feel like there is significant value in a number of protections enshrined in the US constitution (starting with the prohibition on ex post facto laws, as a pretty major one).

* I feel that asset forfeiture is a complete travesty and should never have been allowed under any sort of "limited government" approach.

That sort of thing. Of course then we have to have some debate about which powers governments should or should not have. And revisit that every so often as the situation (society, world power balance, technology) changes. But we should revisit it in a reasonable way, where everyone is clear that we are revisiting it, instead of the government just grabbing more power for itself unilaterally, whether it's through abuse of executive orders, pretending like everything is "interstate commerce", setting up secret courts, doing parallel construction, passing unconstitutional legislation and hoping no one notices, or whatever other things various governments in the US have tried over the last century or two.

I don't know that I've ever seen anything strictly _wrong_ in an article about mathematics, but I've certainly seen articles giving equations without defining what any of the variables are and using using very field-specific notation without defining or referencing it, making them impossible to read unless you know enough about the topic to reverse-engineer what they're talking about.

And I'm not talking something like using a Dirac delta without referencing it; I'm talking about pretty esoteric notation that is really only relevant to the topic of the article.

And of course plenty of the math articles are just not very coherently written (and yes, plenty are quite good).

Because there is a qualitative difference between "limited government" and "no government"?

Just looking at UCLA employees sorted by gross pay is interesting; Tao ends up around #192 on the list. Mostly due to the medical school, plus some people with titles like "INTERCOL ATH HEAD COACH EX" and "INTERCOL ATH COACH AST EX".

The top 3 (and #5) are all "HEAD COACH".

> I have to discount their math abilities.

Do you? The implication that "applied math" doesn't involve math abilities is an interesting one, but doesn't have bearing on reality. There is a good bit of interesting, and quite challenging, applied mathematics going on out there. Including large chunks of what we usually call "computer science".

If I'm a hiring manager hiring for a position where quantitative skills matter, hiring someone who did applied math at a high level is absolutely something I would look for.

Yes, this can be a pain. There are some contracting companies that handle this for you, but that can involve hiring people in other countries as contractors, not employees, having an extra intermediary that handles their payroll and whose incentives in terms of quality service may not align with yours, etc, etc. Plus of course regulations about how long you can have people contracting for you before you have to hire hem as employees and whatnot.

Also, it's not just tax implications. There's also complexity about time off (vacation and leave regulations differ a lot across the globe), conditions under which people can or cannot be fired (or hired), differences in terms of non-salary benefits that must be provided, etc.

A simple example of the hiring/firing thing: under French law, if someone is on maternity leave you can't fire them. You also can't hire someone on an indefinite-term basis to do their job while they are on leave. Hiring someone for the specific term of the leave seems to be OK. See https://www.globalworkplaceinsider.com/2017/05/do-employees-... toward the end, though the whole article is a great illustration of some of the differences in this area between French and US law.

And then there are the fun parts about general regulatory compliance. As a simple example I've run into, the US _requires_ you to collect and report data about the race of your employees (see EEO-1), while the EU _forbids_ collecting this data to start with, last I checked. So you have to have distinct processes for employees in different jurisdictions, your HR database needs to handle these differences in rules, etc.

It depends. If you have to redo the work multiple times because of miscommunications, that can easily eat up salary savings. If you can't hire the best people because of some of the aggravations involved, you may get a worse quality product or it may take a lot more time than you planned, or both.

Again, hiring someone remotely in the way you describe might well make a lot of sense. But it's not nearly the no-brainer you paint it as.

> In the last 10+ years fiscal spending has been half-hearted and anemic

See, this is one of those claims that is interesting to me because I never see anyone proposing what a "non-anemic" level would look like. No matter what gets proposed for fiscal spending, the response is "that won't be enough; we should spend more", without quantifying more.

As a scientific claim, it's unfalsifiable, fundamentally, which makes it hard to work with usefully.

> Austerity does hurt economic growth:

First off, there's a lot of cherry-picking going on. The US did "austerity" things, people claimed the result would be disastrous.... and it wasn't. Europe did austerity, but _also_ very tight central bank policy at the same exact time; results were not great, but it's hard to blame this on just "austerity" given the central bank behavior.

Just to be clear, I am not claiming that austerity _improves_ economic growth, which seems to be what the wikipedia link you post is about. I'm claiming that the specific predictions made about austerity by adherents of specific economics intellectual movements in the last 10-15 years did not in fact match what actually happened.

> but once you hit zero rates, then what more can you do from that angle?

Tons, if you want to, because you are not limited to short-term rates as your policy instrument. Those rates are correlated with monetary policy (in the short term; anti-correlated in the long term), not causative.

Note that at the point when we hit "zero rates" central banks were doing all sorts of monetary tightening (positive interest rates on reserves!, the Fed raising its policy rate 9 times over the course of a few years and consistently missing its inflation target on the low side in the process, etc).

https://en.wikipedia.org/wiki/Negative_interest_on_excess_re... is absolutely a thing you can do to increase V when you hit "zero rates". But more importantly, a large part of monetary policy is expectations management. If people expect you to undershoot your inflation targets, they act accordingly and it actually gets harder to hit the targets.

> Who is writing publicly on this topic, either in weblogs, articles, or Tweets?

https://www.themoneyillusion.com/ has been writing about it for a while now. Less recently, as the idea has gained more widespread traction; a lot more around the 2008 financial crisis and aftermath.

Is the real question here "what is this and why is it supposed to be a good idea" or "who are these people who are making better predictions"?

True, but for the US, for example, energy consumption has been pretty much flat for the last 20 years, and decreasing in per-capita terms. See https://www.bloomberg.com/opinion/articles/2021-01-28/beyond...

It's interesting to see how world trends look as more and more of the world ends up in the position the US was in 20 years ago in terms of economic mix.

Unfortunately, Keynesians also made predictions (like fiscal spending moving rates off the zero bound, austerity hurting economic growth, etc) which did not turn out to be true, due to monetary offset. That is, Keynsians tend to over-focus on fiscal policy and ignore the monetary side, but in practice central banks end up sterilizing most if not all of what's going on the fiscal side, as far as I can tell.

The only group I have seen making predictions over the last 10+ years who seem to be successful are the people suggesting that central banks do NGDP level targeting instead of what they're doing right now.

There are a lot of assumptions going into that implication. My experience with working remotely, on globally distributed teams, for about 20 years now is:

* Synchronous high-bandwidth communication (read: meeting, ideally with a shared whiteboard) is pretty handy sometimes.

* Timezones really do matter. If you want a sync meeting that includes people in Europe, the US, and APAC, someone will have a bad time. If you want quick turnaround on code reviews, having 8-12 hour time zone differences is a significant barrier. Even a "4-hour" difference for meeting time purposes can be a huge problem in this sense: it's 9pm on Thursday where you are (US West coast), but your coworkers' weekend has started (in New Zealand), so any code reviews you ask for on Friday morning you won't see until you get back to work on Monday. And any code reviews they ask for on their Monday morning won't happen until you get back to work on their Tuesday.

* Language matters. Even if everyone involved speaks English to some extent, it's a _lot_ easier to follow some accents than others, especially on an imperfect (read: teleconference) audio feed. Which accents depends on which people.

* Shared context matters, just in terms of understanding requirements. This can be developed, but it takes time. People from more similar backgrounds often require less time to get to mutual understanding around this sort of thing. This needs to be weighed against blinkered thinking and lack of diversity in perspectives, of course.

I love working remotely, and am used to the scheduling issues around this sort of thing, but there are a _lot_ of issues that come with hiring someone halfway across the globe that you don't get with hiring someone also working remotely but in the same city or at least speaking more or less your accent of your language and within an hour or two timezone distance. The fact that the latter works for a particular position does not imply the former will.

I disagree about people being mostly sane about this stuff, for what it's worth. We may have very different filter bubbles....

I'm not sure about the "tech hotspot" angle, but https://www.econlib.org/homeless-camping-in-austin-a-modest-... while somewhat satirical may well have a point about the "why"...

How are you defining "way too much policing"?

Just looking at Western Europe (arbitrarily defined, granted, and leaving out some really small countries) in the list at https://en.wikipedia.org/wiki/List_of_countries_and_dependen..., the US has fewer police per capita than (in increasing per-capita order) Ireland, the Netherlands, Luxembourg, Austria, Scotland, Belgium, Germany, France, Portugal, Italy, Spain. More police than (in decreasing per-capita order) England+Wales, Switzerland, Sweden, Denmark, Iceland, Norway.

The US has about 25% more police per capita than Norway. Germany has 63% more police per capita than the US does.

> even after he discovered

She, fwiw.

It's for two (well, at least two) reasons: one is to prevent things like this. The other is that in generally doing case-insensitive stuff for non-ASCII is language-dependent.

For example, in English "i" and "I" are case-insensitively equal but in Turkish they are not: "i" is case-insensitively equal to "İ" (U+0130) and "I" is case-insensitively equal to "ı" (U+0131).

So if you wanted to do full-unicode case-insensitive things you would need to decide what happens with <DİV> or <dıv> and whether it depends on the language of the page or what. Doing ASCII-only case-insensitivity sidesteps all that, while admittedly privileging English over other languages.

Just to make sure we're clear on the numbers, Parler had ~2 million accounts last I checked. Facebook being 1e6 times bigger, as you claim, would be quite the feat.

I get your larger point, but the actual numbers you picked to illustrate it are just completely divorced from reality.

That's not correct, as I understand it. Section 230 says that you can't be held liable for content on your site that you did not create yourself, even if you generally moderate your content (i.e. are not just a common carrier).

But there are still laws that require you to _remove_ certain content (child porn comes to mind, but also material that infringes copyrights, etc) once you are aware of it.

Concretely, given this sequence of events:

1. You have a web site that allows people to post pictures. You moderate them some.

2. Someone who is not you or one of your employees/contractors posts child porn.

3. Before your moderators notice it, someone else sees it and reports it to the FBI.

4. You remove the child porn (based on normal moderation getting to it, or you getting notified by the person from item 3 or by the FBI).

my understanding is that Section 230 means that there is no legal liability on your part, as long as step 4 actually happens once you become aware of the content.

Since the mute LED is a user-facing security feature, it is in fact a pretty serious bug if users can't rely on it to be correct. Witness the wide variety of incidents over the last year where people ended up in serious trouble because they thought they were muted but weren't.

I feel like of all the examples in the blog post, this was the one with by far the biggest potential for actual harm to people.

> When did speaking truth to power become "bullying"?

It _can_ become bulling when people will arbitrarily claim that the victim of the bullying is the one who "had power" while the mob that got them fired, say, was just "speaking truth". Which maybe they were and maybe they were not.

Not every incident of a bunch of people attacking someone on social media is "bullying", but neither is every incident of such "speaking truth to power", but a lot of the incidents of the former sure do try to pretend to be the latter.

I suspect that you're thinking of a specific set of incidents and that j_walter is thinking of a quite different set of incidents, for what it's worth.

This might be an interesting argument for municipal broadband which occurred to me last night: in that situation the ISP is in fact bound by the First Amendment, which limits its ability to do that sort of thing quite a bit...

> So, for example, it should be illegal to threaten somebody’s personal safety

Even this seemingly-simple example is complicated in practice: who decides whether safety is threatened? I have seen people claim to feel physically unsafe whenever someone raises the question of how specific women's sports should handle participation trans women, and I have seen other people claim to feel physically unsafe when someone suggests raising taxes to fund more public services (immediately reaching for comparisons to the USSR in the 1930s).

Is the test then whether someone feels unsafe as a result of the speech? Or whether a "reasonable person" would feel unsafe? And so on, and so forth.

There's really no good way to win completely here: either some people are going to feel unsafe, or you have to have _quite_ draconian speech restrictions to try to avoid that (and probably still fail). You might be able to do something that works for neurotypical-enough people, but that's the best I can see.

> Free speech refers to government censorship.

No, free speech refers to free speech. Government censorship is one thing that can prevent speech being free, but not the only one.

Or put another way, there is a legal concept of "freedom of speech", rooted in the First Amendment in the US, and there is a somewhat separate moral concept, which can be both narrower and broader than the legal one. There is some pretty widespread disagreement on the exact boundaries of the moral concep.

The legal concept did not create the moral concept. The reason there is a First Amendment is that the value of free speech already existed, and was codified, in one limited domain, in the Constitution.

And I am quite sure that if the moral concept becomes too narrow, with overwhelming support for speech restrictions in the population in general, the legal concept will follow, via the amendment process... Which makes it a priority, for me, to push back on attempts to narrow the moral concept absent great need.

Or to put it another way: There is no legal requirement for a non-government entity to allow various speech in the US. But in many cases I would argue there is a moral requirement, even if they disagree with the content of the speech. Obviously, being a moral requirement, there is a line the entity needs to decide on for itself after which it will no longer support said speech.

The number here ($2.5 billion) seems to be ~20% of Boeing's 2016 EBITDA (earnings before interest, taxes, depreciation, amortization) and ~18% of their 2017 EBITDA. It's ~700% of their 2019 EBITDA (they had a pretty bad year revenue-wise in 2019) and I expect they will have a significantly negative EBITDA for 2020 (it's -$4.6 billion for the 12 months through 2020-09-30).

Looking at historical trends at https://www.macrotrends.net/stocks/charts/BA/boeing/ebitda it looks like it's something like 25% of an eyeballed "average year" for Boeing.

I suspect you're right that this is a little low given what was going on here.

First of all, thank you for your service!

I am absolutely not taking your comments as an assault, and I suspect we largely agree on specific cases.

And specifically, taking down explicit incitement to violent illegal actions by someone already in a position of power does not seem at all unreasonable to me; I can see very little redeeming value in such. Also reasonable is a basic "cooling off" period of "we're not letting you post anymore for a bit, until people calm down".

And I fully agree that there is a long job of getting people to talk to each other, much less be on the same page ahead of us. I hope we succeed at it. Facebook and the like are not going to be helping much with it, unfortunately, because outrage is just so monetizable..

The answer varies. For example, in the US, depending on what you are trying to accomplish it ranges somewhere among "half of the population", "that one guy at the FCC who writes the policy", and "1/3 of either house of Congress or 1/4 of the state legislatures", right?

I did not imply that a newspaper or facebook or twitter have to do anything.

That said, if you want to read things into what I said and actually have this conversation, does my hosting provider have to host my website? Does their ISP have to not cut them off if they do? Under a narrow reading of "free speech" as "stuff protected by the First Amendment", obviously not (unless the ISP is run by a government entity, of course). Under a reading of "free speech" as "the ability of society to not have entities with overwhelming power (governments, but not only them) suppress speech they don't like", the situation becomes more complicated.

And I would argue that a narrow legalistic focus on the First Amendment is not the right way to approach the problem. The First Amendment is a means to an end, not an end to itself. For an 18th century analogy, if a private company has the power to go around destroying all printing presses that don't belong to them and refuses to print views it disagrees with, that is just as much a problem for free speech as if the government were doing it. This was just not a realistic scenario in 18th century North America (unlike, say, contemporaneous India), so wasn't a big concern for the framers of the First Amendment.

Is this a scenario we have to worry about now? It's not clear to me that it's an imminent problem right now, but I can easily see it becoming one in the next 10-15 years. Might be worth thinking about norms (and laws, as needed) now so we're not playing catch-up when it does.

None of that has anything to do with the factually inaccurate claim about the narrow constitutional right that was made in https://news.ycombinator.com/item?id=25673573