It’s pretty typical of corporations, the cult surrounding its leader notwithstanding. Not even just US corporations - the VW emissions scandal was huge, and today they are doing as well as ever. That was a big shakeup; the kind of stuff we are seeing from Tesla feels like business as usual.
Nope - the VW episode was terrible, but they faced large fines and corrected course and it's history. I'm still slightly squeamish about accepting them but they've turned it around and I think I read have just overtaken Tesla in EV sales in Europe (a self-inflicted Musk wound, of course).
I see no course correction from Tesla. Just continuing and utter tripe from it's CEO, team, and Musk-d-riders.
This is an on-going issue for them and, at this point, with no further change? I hope it drives them into the ground (Autopilot, natch).
You can actively criticize VW on the internet without an army of sycophants coming for you. The standard behavior of Tesla stans is that any problem with the vehicle is in fact your fault and only your fault because it would not be possible for Tesla to do something wrong. It is cult-like.
I don't defend it but the specifics never bothered me. They cheated because their cars didn't meet new emissions standards. They were fine by the standards of the year before. So a bureaucracy just declared that a legal level of emissions was now illegal.
In my mind it's like suddenly declaring that blue cars are illegal, and they made a color-shifting car that is blue except when the authorities are looking at it.
It is wrong in the sense that it is normalizion of deviance, however. We live in a society and if we don't like a law or regulation the correct response is to get it legally changed, not to ignore it and cheat.
> They cheated because their cars didn't meet new emissions standards. They were fine by the standards of the year before.
> So a bureaucracy just declared that a legal level of emissions was now illegal.
That is not at all what happened and not how emissions standards are deployed. The EPA's Tier 2 standards were finalized in 2000 to phase in during the 2004-2008 model years [1].
I didn't say you are defending it. I'm saying that "companies do bad things sometimes" is not a full description of the Tesla phenomenon that people take issue with.
If it passes the unit tests I make it write and works for my sample manual cases I absolutely will not spend time reading the implementation details unless and until something comes up. Sometimes garbage makes its way into git but working code is better than no code and the mess can be cleaned up later. If you have correctness at the interface and function level you can get a lot done quickly. Technical debt is going to come out somewhere no matter what you do.
The trick is to not give a fuck. This works great in a lot of apps, which are useless to begin with. It may also be a reasonable strategy in an early-stage startup yet to achieve product-market fit, but your plan has to be to scrap it and rewrite it and we all know how that usually turns out.
This is an excellent point. Sure in an ideal world we should care very much about every line of code committed, but in the real world pushing garbage might be a valid compromise given things like crunch, sales pitches due tomorrow etc.
No, that's a much stronger statement. I'm not talking about ideals. I'm talking about running a business that is mature, growing and going to be around in five years. You could literally kill such a business running it on a pile of AI slop that becomes unmaintainable.
How much of the code do you review in a third party package installed through npm, pip, etc.? How many eyes other than the author’s have ever even looked at that code? I bet the answers have been “none” and “zero” for many HN readers at some point. I’m certainly not saying this is a great practice or the only way to productively use LLMs, just pointing out that we treat many things as a black box that “just works” till it doesn’t, and life somehow continues. LLM output doesn’t need to be an exception.
That's true, however, not so great of an issue because there's a kind of natural selection happening: if the package is popular, other people will eventually read (parts of, at least) the code and catch the most egregious problems. Most packages will have "none" like you said, but they aren't being used by that many people either, so that's ok.
Of course this also applies to hypothetical LLM-generated packages that become popular, but some new issues arise: the verbosity and sometimes baffling architecture choices by LLM will certainly make third-party reviews harder and push up the threshold in terms of popularity needed to obtain third party attention.
I’d make an even broader generalization, which is that all of these attributes are proxies for status. Graduated from a top school? Live in the Bay Area or NYC? Did well in High School (I guess?) These are all status indicators in the mind of someone higher up, and deserving of better pay. If you live in a low cost of living area it’s correspondingly lower status and deserving of lower pay. Other common status indicators include things like age, gender, and race…
Is it possible that people who are good at acquiring status markers are also good at other things? No, it’s the world that’s wrong - there is no reason a Harvard grad in NYC should have a higher chance of being paid six figures than a University of Missouri-Kansas City grad in Missouri.
I certainly didn’t mean to imply otherwise. I do think that hiring solely or even principally based on superficial status indicators can lead to systemic problems, especially when those indicators may not be good reciprocal proxies for relevant skills.
Crypto moves the problem from payment providers like Visa to central exchanges like Coinbase. Until you have a completely decentralized ecosystem built around crypto, you run into trouble when offramping to fiat. If I recall, backpage accepted bitcoin when Visa dropped them, but it was way too much hassle to be useful. If you could pay rent and utilities and buy food using some sufficiently decentralized token, crypto may become a viable alternative.
The principle should be that it shifts the problem to payment providers who can be switched out for other payment providers seamlessly. The providers are motivated to behave ethically because you have the option of going elsewhere.
Paying with crypto is still not very usable but you can still do it directly which limits the degree of extortion that can be applied. I think it will get better as it ceases to be 'interesting' and people develop tools that just work rather than try to revolutionize your life.
The interesting thing about cryptocurrencies is using them directly, i.e. when users have their own wallets under their full control. Then it's magical when you make a transaction to somebody and think that nobody is censoring, filtering, moderating or rejecting it in any way. Oh and no PII either.
Note that pornography is not banned here in Texas at least. You just have to provide age verification, and PH elected not to participate in that process. It doesn’t seem like that wild a thing at face value.
It would take a lot of power to send even a radio signal that could be picked out from the noise at a few light years. Add a requirement for that signal to be more or less continuous over geologic timescales - we’ve only been able to emit and detect these for ~100 years - and my personal surprise diminishes rapidly. Huge distances in time and space with human-level technology make detection highly unlikely.
Yes, and I would add my favorite hypothesis to the paradox, an anthropocentric assumption theory of self importance... or let's call it an anthropocentric bias:
Humans tend to define intelligence, life, and communication based on our own structure -carbon-based biology, electromagnetic signaling, language, symbolic thought, etc. This narrows the scope of our search.
We assume other civilizations want to communicate, would use similar media (radio, light, mathematics), and would send signals we could interpret. This ignores other potential modalities (quantum, neutrino, gravitational, exotic matter, etc.) or entirely non-signal-based forms of interaction.
We may not even recognize signs of intelligent activity if they don't resemble our expectations, ie entire civilizations could exist in forms of computation or energy we can’t perceive.
We assume ET intelligences are aligned with our timeframe or curiosity. Maybe they don’t care to communicate, see us as trivial, or operate on million-year attention spans.
It may reflect less the silence of the cosmos and more the limits of our understanding, especially the assumption that we're capable of detecting or interpreting intelligence beyond Earth. A epistemic humility, or rather our lack of it.
Nobody would be communicating with neutrinos or gravitational energy. EM radiation is way easier to emit and detect, and at cosmic distances they all scale exactly identically (inverse square law). The other things you mentioned are mostly sci-fi inventions and there’s nothing in known or unknown physics that would hint towards them being plausible communication media.
It’s not about being shortsighted, it’s about everyone being constrained by the same laws of physics. Our models, however imperfect, are still unreasonably good.
The counter argument is that even if civilizations exist with all the properties you described, given the vastness of space, there should be another civilization that pattern matches to us.
Sure! All the hypothesis in the Fermi paradox deal ultimately with calibrating our expectations of making contact, not with denying the existence of "STEM-enabled" species like us yearning for an alien encounter.
There's epistemic humility, then there's indulging in unfalsifiable fantasies in the name of not ruling anything out.
> Humans tend to define intelligence, life, and communication based on our own structure -carbon-based biology, electromagnetic signaling, language, symbolic thought, etc.
I would posit that none of these properties are coincidences, and are in fact likely to evolve convergently in most if not all circumstances hospitable to life. In particular I very much expect ET life to be carbon based; I don't believe there's a true viable alternative outside scifi (hint: silicon ain't it).
> entire civilizations could exist in forms of computation or energy we can’t perceive.
Could they? Really? There aren't that many gaps in the Standard Model. The aliens could be made of dark matter, I guess, and remain forever undetectable, but that's not to far off believing in invisible fairy kingdoms. And it still wouldn't explain why the baryonic sector is so devoid of detectable life. Ethereal undetectable aliens don't mean regular ones can't also exist.
> Maybe they don’t care to communicate, see us as trivial, or operate on million-year attention spans.
This one I'll grant (sort of: what's the evolutionary path toward such entities arising?), but it's still weird that we haven't seen any sign at all of them. These entities live on million-year timescales but have no visible effect on their surroundings? Why?
And more importantly, why is that the only thing that happens? Because if it isn't the only thing, then the question remains of why can't we see anything else?
Radio signals aren't the only sign. I'd really love to see some sign of megastructure engineering, but even detecting O2 in an extraterrestrial atmosphere would be huge.
This has never stopped working for me - I interview every year or so as a matter of course.
My experience helping others is that in a tightening labor market they are experiencing "small fish in a big pond" syndrome mentioned above.
Compensation for the bottom 50-75% of talent has dropped significantly since 2021* - if you are in that band and chasing the same comp you're dealing with the "small fish in a big pond" situation. If these people drop their compensation expectations back to 2017 levels many of them will find jobs much more plentiful - though the very bottom of the pool is being driven out of the industry entirely.
* this estimate pulled from conversations with engineering leaders and recruiters at large SV tech companies
What's up with the typescript rant? Don't we have decades of research about the improvements in productivity and correctness brought by static type checking? It starts out strong but stuff like this really detracts from the overall point the article is trying to make, and it stands out as an incorrect and unjustified (and unjustifiable) point.
The whole thing is basically you can't know anything, and personal experience can't be trusted. So we are all helpless until "Science" tells us what to do. Until then sit on your hands.
Where is the proof that Javascript is a better language than Typescript? How do you know if you should be writing in Java/Python/C#/Rust/etc? Probably should wait to create your startup lest you fall into a psychological trap. That is the ultimate conclusion of this article.
It's ok to learn and experiment with things, and to build up your own understanding of the world based on your lived experiences. You need to be open minded and reevaluate your positions as more formalized understandings become available, but to say it's too dangerous to use AI because science hasn't tested anything is absurd.
> Where is the proof that Javascript is a better language than Typescript?
This is an interesting question really. It feels like it would be really hard to do a study on that. I guess the strength of TS would show up mainly as program complexity grows such that you can't compare toy problems in student exams or what ever.
> The whole thing is basically you can't know anything....
Well that position is hopelessly circular, filled with sophistry and fallacy.
Descartes proved that one wrong in the 1600s. "Cogito, ergo sum,"
> Where is the proof that Javascript is a better language than Typescript?
Any 'best' question relies upon an answer to the question of for what objective purpose. Not providing that answer, is the same as dissembling circularly, and any conclusion based on that is more likely to be false than true. False beliefs in this manner are by definition, delusion.
As far as I know we don't have decades of research about the improvements in productivity and correctness brought by static type systems.
We have one study on test driven development. Another study that attempted to reproduce the results but found flaws in the original. Nothing conclusive.
The field of empirical research in software development practices is... woefully underfunded and incomplete. I think all we can say is, "more data needed."
hwayne did a talk on this [0].
If you ever try to read the literature it's spartan. We certainly haven't improved enough in recent years to make conclusions about the productivity of LLM-based coding tools. We have a study on CoPilot by Microsoft employees who studied Microsoft employees using it (Microsoft owns and develops CoPilot). There's another study that suggests CoPilot increases error rates in code bases by 41%.
What the author is getting at is that you can't rely on personal anecdotes and blog posts and social media influencers to understand the effects of AI on productivity.
If we want to know how it affects productivity we need to fund more and better studies.
> As far as I know we don't have decades of research about the improvements in productivity and correctness brought by static type systems.
I seem to recall at least some research dating back to the 90s on the topic, which showed how much better (by some metric I can't remember) Ada was wrt most other languages of the time.
"Ada is more productive than contemporary languages" is a much narrower statement than the implied "static typing is always more productive and correct than dynamic typing"
> Don't we have decades of research about the improvements in productivity and correctness brought by static type checking?
Yes, we have decades of such research, and the aggregate result of all those studies is that no productivity gain can be significantly demonstrated for static over dynamic, and vice-versa.
Not sure what result you are referring to but in my experience, many of the academic research papers use “students” as test subjects. This is especially fucked up when you want to get Software Engineering results. Outside Google et al where you can get corporate sanctioned software engineering data at scale, I would be wary most academic results in the area could be garbage.
> the debate about types, TypeScript, and web development is, for example, largely anecdotal gossip not backed by much in terms of structured research
I'd be interested in the type of structured research the author is interested in. Could it also be researched whether Go or PHP is better for web development? In some sense, I guess.
Both are probably more efficient than writing Apache extensions in assembler, but who knows?
It’s ironic as static typing exists precisely because developers don’t trust their instincts?
It’s a formal acknowledgement that humans make mistakes, implicit assumptions are dangerous and that code should be validated before it runs. That’s literally the whole point, and if developers by type were YOLO I’ll push it anyway, TS wouldn’t have got anywhere near the traction it has. Static typing is a monument to distrust.
You mean like the autocomplete server/LSP shows you the results faster if you're doing TS than JS? Sounds like a weird editor quirk to bring up if so. In neovim the autocomplete is as fast in JS as any other language, not sure why it wouldn't be like that for you.
No, it's because TS adds information. Pure JS doesn't know what variable "foo" is (is it a potato? is it an array? can it be null?), but if you assign it a type like Map<string> then now it knows what kind of data it can hold, what you can do with it, etc. Now you know you can do foo.get() and your IDE will tell you as soon as you've typed `foo.`, but if you used pure JS, lol how is it supposed to have any idea what methods exist?
It lets you chain data from one task to another based on knowing what possible and available given the types. That information just straight up doesn't exist in JS code, unless you write JSDoc comments all over with that information as static analysis hints. But why not just embed that information in the language itself instead of tacking it at the top of every chunk of code, much more elegant and allows way more power.
Even if you used JSDoc, you still need that static analysis tool to exist, and hey, that's exactly what TypeScript is. It's both a language definition for augmenting JS with types, but it's also a transpiler which takes TS code and spits out JS code, _and_ it will tell you if there's problems with the types. But since that engine exists, you can plug it into LSP and so on.
Remember, LSP exists _because_ of TypeScript. VSCode exists because of TypeScript. The team that works on VSCode and came up with LSP did it because they wanted TypeScript to integrate with VSCode and they had the foresight to build something generalized that could be used for ANY language.
> Don't we have decades of research about the improvements in productivity and correctness brought by static type checking?
correct me if I'm wrong those studies would not have looked at TypeScript itself, which for all I know could be complete garbage designed to lock you into MSFT products.
Just because they weren’t born into the owner class (or “capital class”) doesn’t mean they didn’t work their way into it. That’s kind of the American dream.
You can only become wealthy later in life at which point you can’t advantage your past self. Thus new money receives fewer benefits than old money from the exact same policies.
Further having 100m at 40 doesn’t suddenly bring the kind of social connections that going to the right schools and the right parties would. At the extremes, the average lottery winner is surrounded by people asking for help, the average Fortune 500 CEO’s social circle aren’t. So if they suddenly fall on hard times the lottery winner is stuck but that CEO may very well claw their way back.
It’s still possible for poor people to succeed and 3rd+ generation wealth to fail, but the odds are wildly different.
reply