In addition to covering the IPO in general last week, Matt Levine also wrote about this specific question Tuesday[1]:
> Historically index providers were in the business of making these sorts of quality decisions, so that index funds were not forced to buy stocks they didn’t like.
> These rules create some tension between the idea that an index is a list of all the stocks and the idea that an index is a list of all the good stocks. Historically, it didn’t matter all that much: The point of the stock market is to tell you which stocks are good, so a company with a high stock valuation should be a very good company, so it should get a high weighting in both the Index of Good Companies and the Index of All the Companies.
> But SpaceX — and also maybe OpenAI and Anthropic in their coming IPOs — will probably break that link. SpaceX will probably (1) do all sorts of stuff that index funds hate and that index providers have specifically tried to exclude and also (2) be gigantic, because the market loves it.
The original decision is interesting because at first it seems very stupid (as acknowledged right there in the article). It's a more expensive way to do the same thing. But man, what a sales pitch, not only for their own customers but also to employees. The feeling is that the company values its people and is willing to really depend on them, and look, it actually paid off when they did that.
I think it's common to claim to care about the people without really depending on them for much (like with perks) or to depend on the work but treat people badly, and doing both is hard.
Yeah, I mean it's obviously meant to be a marketing pitch but it's not a very good one.
> The hardest computational problems are not waiting for faster chips – they are waiting for machines that compute in a fundamentally different way.
Surely they don't actually believe that, right? Like you say the benefits must be limited to specific shapes of problems (not all of "the hardest" ones), and the whole history of computing is about how faster chips is an excellent answer to difficult computational problems.
> and the whole history of computing is about how faster chips is an excellent answer to difficult computational problems.
I don't really disagree, and I am definitely not taking their marketing pitch seriously. Yet, you could look at the same computation history and interpret it as an economically constrained hill-climbing around an idea that was simple enough to work reliably (von Neumann architecture) and that worked and scaled so well that we were rarely forced or desperate enough to move conceptually far away from it.
Sufficiently general digital computers can simulate other computational models, so I think 'faster' is ultimately the end game, but for some classes of computation, as you also noted, we may need to go for analog hardware, (maybe) quantum devices, optical interconnects, and so on.
Bret Victor has a talk about this, more or less: [0]
That's interesting, thanks. I only read the abstract so far but was immediately reminded of this recent HN submission[1] and the whole thing that certain ideas go together, and so they are adopted together, but the resulting bundle of ideas might be poorly suited to certain problems.
Discussion of stats models is always complicated by the fact that a lot of people will read "30%" as a "no" prediction and claim your model is wrong if the thing happens. On the one hand, one strategy is to "hide" the numbers a bit behind a blaring headline that says "we are not sure!!" It's a bit of an art to decide when to be "sure" or not. On the other hand, in research for example you can just say screw it, I care if the correct people are correct, not if a bunch of wrong people are wrong.
I feel like the correct strategy for 538 when it was actually niche was to be precise, but then it went viral and maybe should've hit the IDK button much harder and more often after that.
The real caveat is that 538 was a Monte Carlo model, and is only as good as its inputs. "Here's what the current spread in polling numbers is *given our model and the current polling and their reported uncertainties.*" Polling uncertainties are themselves computed under certain models, and those models are subject to errors. I don't think 538 hid this, but it's a difficult caveat for people to reason about because the sorts of modeling errors that have the most influence usually represent "unknown unknowns".
Building a model for predicting the ultimate winner of a US presidential election is particularly difficult, because you are dealing with noisy input data and nonlinear effects, i.e. just a few thousand votes in a few key states can completely flip the outcome. If you then have poorly calibrated polls with a large margin of error, there is really nothing much you can do.
On the other hand, it does raise the question how valuable the 538 models for something like this really are if the outcome is a coin flip anyway.
Exactly, and correlated errors, where a polling error in one state predicts similar errors across the board.
I disagree that it's all pointless though. Most basically it's smart for campaigns to have a good model and let that inform strategy where appropriate. Since the president is a big deal other people's decisions are also impacted, and in the long run it pays to have good predictions of those chances. Also, the outcome sometimes is fairly certain and that isn't always easy to see.
I agree, it's far from pointless. The 538 model is arguably close to the best you can do considering how difficult the task is, but it's important to understand it as purely a reflection of the polling data (and 538's reliability scores for polls), and that polling data is inherently flawed. After all, there are only 2 ways to perform a perfectly accurate poll: either know the outcome a priori, or run the election. We shouldn't be too surprised when models like 538 fail to correctly predict the outcome, because that's not what they represent. It's an analytical tool for understanding the current state of polling.
> Discussion of stats models is always complicated by the fact that a lot of people will read "30%" as a "no" prediction and claim your model is wrong if the thing happens.
I've even heard things like "70% chance of Hillary winning means she gets 70% of the votes!" (and tangentially, my far-too-long argument with someone on Reddit who insisted "there is no way in hell 50% of the people in this town make above the median income"...)
People don't like seeing a 95% chance of winning and then losing. The game tweaks the odds, so certain thresholds become gimmes (something like "if the displayed odds are better than 75%, treat them as 100%").
Seems weird that it would piss you off, if you were really that invested in the cold hard stats you'd know that if it was fair rng you could still have been the 1 in 100,000 player that got lucky on 75% 40 times in a row.
Fire Emblem does something complex with averaging random numbers to do the same thing - a 95% chance to hit becomes 99.5, and the reverse for low percentages.
The thing jumping out at me is these really are mini businesses (even though they are bad). Combine it with the main idea in "Emacsification of Software" (from recent HN front page [1]) and I guess you end up with lots of nerds running their own customized mini businesses?
It's sorta wild to think about. Am I the owner of the custom radio station my AI agent made, and does that mean I get paid for listening to the ads?
Maybe the cost of computing and running the station means it still needs a decent following to break even, not sure how the numbers work out.
I think there are many ways someone with his lack of expertise can still be valuable, including:
- Making connections to other subjects that an expert would miss. The hall of fame of sigmoid predictions is just excellent, I already know I'm going to be reminded of it some time in the future. Very entertaining way to get the point across.
- Writing about tricky concepts in a very accessible and elegant way, which experts are notoriously bad at doing themselves - they are often optimizing for other specialists.
- Being able to write with an air of speculation and experimentation with ideas that experts and institutions often can't afford. Experts have to maintain their track record; Scott Alexander can say "lol just double the timeline"
you do you, I don't come here for superficially informed-looking articles written by people who are in fact not experts, informed or educated, I come here for the real deal
it doesn't help that sCotT aLexAndEr is also as close as you can come to the modern dressed up version of a eugenicist (again, not based on any actual expertise)
> as close as you can come to the modern dressed up version of a eugenicist
Their writing about genetic determinism is a turnoff to me too. But this essay is about a different topic, and a piece of writing by a writer who is known for writing substantively about a variety of topics should be evaluated on its own terms.
As a fun exercise replace AI with "junior" and "junior" with "mid-level." It holds up pretty well, as a manager you have responsibility for the work your team does and "make everyone put in more hours for no reason" is dumb. Maybe it comes across a bit neglecting of the "juniors" (in particular, it doesn't show any desire for figuring out ways for AI/"the juniors" to grow their responsibilities in a sustainable way).
Imagine reading that version as someone who doesn't know how big companies work. "But then they'll just fire all the mid-level managers, since they don't do any of the actual work!" Haha, boy would you be wrong.
Wow, this really changes how I think about working with software and with LLMs. Sharing ideas and amateur remixing and setting up something weird for you and your friends is so much easier now. Things you had to have lots of time and expertise to do before are just widely accessible now.
It's kinda fascinating how dominant LaTeX is, how nice its output is, how respected Knuth is as a computer scientist, and at the same time how totally awful it feels to use it. Hard to figure out how it can be so good and so bad at once.
In particular it's interesting how people seem to think TeX itself is actually quite nice to use but its popularity and LaTeX packages created a huge mess of a system.
Well -- TeX is "80s good". We've gotten better at designing ergonomic software since and it really doesn't meet the modern standard. But it's good enough for most people, and sufficiently hard to replace, that it has stuck around.
Added to that, academics specifically are more willing to suffer old crufty stuff than software engineers tend to be. After all their job is to absorb fields of material whether good or bad, and the technology tends to be lagging behind the bleeding edge in many subfields anyway so TeX doesn't even necessarily stand out.
Bingo. Compared to troff and what preceded, TeX was amazing just in its usage. But its real value was in the quality of its typesetting. Knuth put a lot of effort into the beauty and historical correctness of the output, so much so that it was solving optimization problems to calculate line breaks. MS Word still can't break a line properly in 2026.
If TeX is “80s good”, Typst might be “90s” good, being generous.
Celebrating batch-mode typesetting in 2026 feels like some weird cyberpunk fixation.
Programmable like Emacs (but via Scheme), interfaced with major Computer Algebra Systems, tree-structured documents that are live-queryable and modifiable, and typesetting that rivals TeX without using TeX - TeXmacs provides all that, and much more (https://www.texmacs.org/tmweb/home/videos.en.html)
There was a point in the 1990's where microsoft word wasn't truly WYSIWYG. IIRC it was like an infinite page and the line breaks and page breaks were "estimates"
Further many docs from that era are plagued with abandonware.
TeX did one thing well for an era when often the only interface to the machine was over a Xyplex terminal server connecting to a tty at 9600 baud.
You're linking to posts from 15 and from 18 years ago. And the post from 2011 is about how Donald Knuth wrote TeX (not LaTeX) in the early 1980s. While TeX and LaTeX have fundamental design flaws, it is much less awful to use them these days, with a rich selection of rather robust packages available, that vastly reduce the need to go into hard-core LaTeX programming yourself.
I won't lie: It takes getting used to and you need to learn a lot if you want to achieve fancy complex typesetting effects. But - it's not half as inconvenient as it once was.
part of the challenge is the inherent irreducible complexity of the domain. "Make text look good on page" leaves lots of details unspecified.
another part is many people built their own solution to their own corner of this domain, and not all of them had the deep appreciation for how the rest of the TeX system works.
I hear similar complaints about "Make web page look good", which is popular but also a huge mess of a system.
> "Make text look good on page" leaves lots of details unspecified.
Even just a sane layout renderer is incredibly hard. A decade ago I wrote a bespoke DNA sequence typesetter (in svg) and I had claude build an extension, for whatever reason it chose to build it from scratch instead of using the components I had built, and it did everything wrong.
to be fair to knuth, he had nothing to do with latex. it's conceivable that one could start over from plain tex and build up a different high level system. (then again perhaps some of the brittleness of latex comes from unavoidable issues with the tex layer; lamport is a very respected computer scientist too!)
The dichotomy comes from conflating the TeX syntax with tex macro system, both use backslash.
The backslash based syntax allows for some really powerful typesetting which is far above anything that exists today. At the same time, the use of backslash-based langauge right to the bottom in terms of macros is what is causing the frustration.
Typst kind of solves that by having backslash based syntax implemented in Rust.
> Historically index providers were in the business of making these sorts of quality decisions, so that index funds were not forced to buy stocks they didn’t like.
> These rules create some tension between the idea that an index is a list of all the stocks and the idea that an index is a list of all the good stocks. Historically, it didn’t matter all that much: The point of the stock market is to tell you which stocks are good, so a company with a high stock valuation should be a very good company, so it should get a high weighting in both the Index of Good Companies and the Index of All the Companies.
> But SpaceX — and also maybe OpenAI and Anthropic in their coming IPOs — will probably break that link. SpaceX will probably (1) do all sorts of stuff that index funds hate and that index providers have specifically tried to exclude and also (2) be gigantic, because the market loves it.
[1]: https://www.bloomberg.com/opinion/newsletters/2026-05-26/ind...
reply