krick

There is no such person as "Mozilla". There is Kiki, a "Support Community Manager", probably a relatively low-level worker (but it doesn't matter much if she is actually has some weight in the organization). So, you are Kiki. You just saw that message. What do you do now? Just ignore it? Do not respond anything and immediately call the CEO and try to convince him/her that he/she must order to disable that auto-translation bot, without even trying to gather more information? No onuses and stuff, what are your actions, exactly?

Because a lot of people in this thread are whin… ahem, expressing their discontent with Mozilla, as we all usually do, but I've yet to see anybody to propose anything realistic at all, let alone better than ask an offended community member for a call and at least to try to talk it through and establish what could be some actionable steps to remedy the situation.

(Unlike GP) I don't actually have a problem with your assumptions. They seem likely to me. But I still have a problem with the whole sentiment of, uh, people on your side of the discussion.

Let's just assume it is how you say it is. (The only assumption I am not willing to make is that people at Mozilla are already convinced it was a bad idea after all.) What in your opinion would be the right move now, after they rolled this bullshit auto-translator out and pissed off a lot of people in the community, including a major contributor for the last 20 years? Surely they could just ignore him and go on with this auto-translation initiative (BTW, thay don't even have to worry about whatever he wants to "prohibit" to do with his translations, because he waived off his rights by posting them). Would it be better than trying to set up a call and discuss things, try to find some compromise, gather a number of recommendations she may then pass onto people working on the auto-translator initiative (because surely this Kiki person, whoever she is, is not the sole person responsible for this and cannot magically just fix the situation)?

I can imagine what you mean, but since I am not in Japan, it would be interesting why you feel that way.

I still don't quite understand where ChatGPT and its pals learned this. Sure, all these PR copywriters are notoriously bad at writing, but still, I don't think I often met all this crap in many texts before. I mean, if I did, I wouldn't be noticing it now as that ChatGPT style. So why does it write like that? Is it even how Anthropic models write as well (never used them)? Is it some OpenAI RL artifact, or is it something deeper, something about the language itself?

I cannot even always quite formulate, what irks me about its output so much. Except for that "it's not X, it's Y" pattern. For non-English it may be easier, because it really just cannot use idioms properly, it's super irritating. But I wouldn't say it doesn't know English. Yet it somehow always manages to write in uncannily bad style.

So… what's changing? Profiles have been there for decade(s), is it just an alternative UI for folks who were unable to run firefox -p, or is there anything more substantial?

I still cannot understand how it came about that everyone adopted systemd, despite it being a showcase of the most selfish, reckless and irresponsible maintainership ever. And it's not like it's something new, it's been like that for what, 10, 15 years now?

FWIW, it says the same when I asked deepseek that question. And while I cannot prove otherwise (I didn't try to specifically do that), I am under very strong impression that past chats influence the future ones. This could be some kind of cognitive bias, but there were some very suspicious coincidences.

I still somehow haven't tried Claude Chat, and while I wouldn't assume it lies about if it remembers anything, I wouldn't just trust whatever these things say about themselves either.

> ChatGPT offers a “memory” personalization setting ... OpenAI are at pains to point out that this function can be turned off at any time, and that individual memories can be deleted

Uh… Isn't it just irrelevant (to the point of such remarks being actually misleading) anymore? AFAIK, it's been a couple of months already since OpenAI began storing all your conversations (because of that court order), whether you "delete" them or not, so while you can technically disable "memory" setting, it only means it won't be able to use your past responses to help you. But it surely would help anybody with, let's say, elevated access. Granted, the threat model in the post assumes that the author is only worried about what the user of the account can learn about other users of the account, and that he trusts OpenAI itself. But why would OpenAI be "at pains to point out that this function can be turned off" then?

Genuinely curious: is it even still relevant today? I've got the impression that there were a lot of these elaborate techniques and algorithms before around 2016, some of which I even learned, which subsequently were basically just replaced by some single NN-model trained somewhere in Facebook, which you maybe need to fine-tune to your specific task. So it's all got boring, and learning them today is akin to learning abacus or finding antiderivatives by hand at best.

Exactly. This is even more annoying when it isn't exactly a hash, but some gibberish you cannot really make sense of, which does have a numeric section in them: like a user ID, or unix time, or who knows what else it could be, but you are trying to visually find a file abcd89764237 somewhere after abcd683426834, and it isn't evident why you cannot, unit you notice that the latter has more digits in its "ID" for some reason.

I think "disastrous" is a bit too strong of a word, but I don't see any "real" reasons mentioned here why it won't be. Sure, there are "cheap" phones that are (almost always) Android, it's also probably true that "most people" use those and wouldn't switch no matter what. But there are also Pixel phones, Samsung Galaxy phones, you know, what people call "flagships". Why people buy these? It's been a long time since they stopped being competitively cheaper they Apple. Even flagship Huawei phones now cost the same as an iPhone. Who buys them? Well, I did. Solely because I couldn't install software I want on iPhone. If I truly won't be able to install what I want on my Android phone — I don't know yet, how I'll deal with that issue (surely I'll figure it out) but I promise you — I'll buy an iPhone for the first time in my life, if only to say "fuck you" to Android. And I urge you to do the same. Vote with your wallet.

But wouldn't you run whatever you are writing in Docker container on dev anyway? I always did just that, as long as I use Docker as a deployment tool for a project at all. And, in fact, even sometimes when I didn't, since it's much easier and cleaner to keep, say, multiple php or redis versions in containers, than on your PC. And it's only maybe a year since it became actually effortless to maintain all this version-zoo locally for Python, thanks to uv. In fact, even when it's something like Go, so the deployable is a compiled binary, I'd usually wrap the build process into a Docker container for dev anyway.

Exactly. I wanted to also point this out in the relation of the author's desire to put all build commands in `just` configuration file. It sounds to me like a desire to use some another "slick and shiny tool" (which `just` is when compared to `make`), but what's the point exactly? The build-process will still be container-dependent and may or may not work outside of the container, and you don't get the benefit of Docker caching anymore.

Recently I somehow wasn't using LLMs locally and relied mostly on ChatGPT for casual tasks. I think it was a little less than a year since I played with ollama, and I remember that my impression was that all recent popular models definitely aren't "uncensored" in a sense that some older modification of llama2 I used was, and all suck for prose-related tasks anyway. In fact, nothing but ChatGPT models seemed good enough for writing, but, of course, they refuse to talk about pretty much anything. Even DeepSeek is not great at writing, and it it much bigger than anything I ever ran locally.

So, are there even good uncensored models now? Are they, like, really uncensored?

A model changing its opinion on the first request may sound more flattering to you, but is much less trustworthy for anybody sane. With a more stubborn model I at I have to worry less that I give off what I think about a subject via subtle phrasing. Other than that, it's hard to say anything about your scenario without more information. Maybe it gave you the right information and you failed to understand it, maybe it was wrong and then it's no big news, because LLMs are not this magic thing that always gives you right answers, you know.

Can somebody explain what's going on here? It seems I'm missing some important piece of background info. Why don't they just add -J flag for everyone who wants to output JSON? Oh, wait, tree already has -J flag to output JSON. So WTF are they doing here?

I am especially confused by this:

> Surely, nothing will happen if I just assume that the existence of a specific file descriptor implies something, as nobody is crazy or stupid enough to hardcode such a thing?

Wait, what? But "you" (tree authors) just hardcoded such a thing. Do "you" have some special permission to do this nonsense?

I haven't used Gemini Pro, but what you've pasted here is the most honest and sensible self-evaluation I've seen from an LLM. Looks great.

It would be great if it was impractical, but I still think it has to be made illegal. I think it is not only "bad" because it's bad for Valve of whoever wants to sell or buy these games, this is "bad" on its own, because it is simply another elaborate form of blackmail. It is as close as it gets to violence without actually beating one up: you are forced to do something not because you are required by law or by nature, but because I want you do that, and I have power over you. This isn't a unique situation in that sense, of course, this pattern appears a lot in life. This one is unique only in that sense that it targets a very narrow bottleneck. If you have power to influence a payment system, and it has legal right to choose to write regulations like that, then you have power over virtually anyone and everyone. Lobbying a policy clause for Mastercard is basically as good (or maybe even better) than lobbying a law. Which, yes, is a problem on its own, but this problem is rather "an unfortunate situation" than a malicious act.

Also, on the matter of the latter problem — fixing that is much easier said than done. Cryptocurrencies in a way were an attempt to fix that, but governments around the world do not want this problem to be fixed. And of course they don't. The field is highly regulated, because, as I've said, having control over the payment network is not that very much different from having a physical army to physically beat you up.

As an aside, I was especially surprised by:

> We see this at work with hosting.

Do we, though? It's another topic of course, but I actually share the sentiment that the current trend is the opposite one: good old days of the Free Web are gone, and the reality is that in the days of Gmail (and other major mail providers), Cloudflare (and other major CDNs) the Internet is inherently not a decentralized structure anymore. It takes a few powerful friends to reach an agreement with each other, and everyone else has to follow.

So, anyway, what I had in mind was exactly that:

> to regulate payment processors like common carriers and require them to serve everyone equally regardless

But, as I've said, I don't quite see how to draw the line here. After all, it would be somewhat unfair to payment networks to trip them from making any choices. Both because formally they are just a business, not a commodity (which may actually be the root of the problem, I think), and they should have some right to choose how they want to operate; and also because different customers and products objectively carry different risks. So they have to be able to produce some policies, these policies just have to… to be restricted to what's necessary somehow.

I wonder how this type of pressuring can be made illegal. To be clear: I certainly think this has to be made illegal, and even attempt to coerce a payment network to force a business to do something should be a serious crime. If the product is not illegal, you can sell it, either on your platform, or on a third party platform (Steam), if the platfrom is ok with selling your product. It's arguable, but perfectly reasonable to say that Steam can also choose to not sell it if it doesn't want to. But then, of course, it doesn't seem like a huge leap to say that a payment network can choose to not handle transactions for some type of business if it doesn't want to. Sure, you can appeal to it being a de-facto monopoly, but isn't Steam a de-facto monopoly as well? I mean, I have some trouble formally drawing a line here, yet it clearly seems not right to me, that a payment network can choose at all if they want to handle this or not.

And, as I said, attempt to apply pressure on a payment network, in order to apply pressure on its customer, in order to apply pressure on their customers... well, I think it's pretty obvious why this is a problem, and that things are not supposed to work this way.

Oh great. Now I suddenly feel the urge to acquire a physical oddly shaped lens or mirror.

Instagram doesn't actually work, right? All frontends are down, and it doesn't seem to work locally either.

How big is English in "English grammar checker"? Is it plausible to add other languages to it, or the underlying framework is so English-specific that it doesn't make sense to even bother building something else than English grammar checker upon it?

Somehow the existing answers don't satisfy me, so here's my attempt. The essence of it is really simple.

The axiom is an obviously true statement: if you have a bag of beans, you can somehow take one bean out of it, without specifying, how do you choose the exact bean. Obvious, right? And that's really it, informally this is the axiom of choice: we are stating that we can somehow always do that, even if there are infinitely many beans and infinitely many bags, and the result of your work may be a collection of infinitely many beans.

Now, what's the "problem"? If you look closer, what I've just said is equivalent to saying we can well-order[0] any set of elements, which must make you uncomfortable: you may be ok with the idea that in principle you can order infinitely many particles of sand (after all, there are just ℕ of them), but how the fuck do you order water (assuming it's like ℝ — there are no molecules and you can divide every drop infinitely many times)?

This is both why we have it — ℝ seems like a useful concept so far; and the source of all notorious "paradoxes" related to it — if you can somehow order water, you may as well be able to reorder details of a sphere in a way to construct 2 spheres of the same size.

[0] https://en.wikipedia.org/wiki/Well-ordering_theorem

That's a really good take, and I personally missed that his departure was pre-planned all along (you are the first I saw to mention 130 days). But, again. "So, thank you, Elon, as you are leaving your role anyway, how about making a little performance for the public? Be my friend, post on Twitter that I didn't release Epstein files because I'm in them…"

…Really?

That's what I usually think too. Even if just to be cautious: people alluding to "Hanlon's razor" (as if it's a real thing) are basically declaring themselves the smartest in the room, so by another well-known eponymous effect they are usually the dumbest in the room. Usually the worst suspicions are confirmed later.

This time, though, I'm running with the crowd. I think this is just too much. I mean, come on, screaming on Twitter that Trump didn't release Epstein files, because he is in them? Sure, it doesn't hurt him, it's no news nor a real accusation, but I'm pretty sure Trump didn't want that to be posted. The whole thing doesn't look nice for anybody, it doesn't help anybody. No, I really think Musk has become totally insane this time, or/and is drugged out. The left hand still may be doing something, but that's taking the opportunity, not making this all up for the sake of distraction.

It's also impossible to turn off in my experience. I have like 5 lines in my ChatGPT profile to tell it to fucking cut off any attempts to validate what I'm saying and all other patronizing behavior. It doesn't give a fuck, stupid shit will tell me that "you are right to question" blah-blah anyway.

It's not like I didn't call Musk crazy before, but I think now he is truly, really nuts. Like, clinically insane. This isn't about the Dragon, of course, but about the whole thing. Very curious to see how it ends and if he is as untouchable as he clearly thinks he is.

In all fairness, the essence of it doesn't have to do anything with copyright. "Pro-copyright" is old news. Everyone knows these companies shit on copyright, but so do users, and the only reason why users sometimes support the "evil pro-copyright shills" narrative is because we are bitter that Facebook and OpenAI can get away with that, while common peasants are constantly under the risk of being fucked up for life. The news is big news only because of "anti-ChatGPT" part, and everyone is a user of ChatGPT now (even though 50% of them hate it). Moreover, it's only big news because the users are directly concerned: if OpenAI would have to pay big fine and continue business as usual, the comments would largely be schadenfreude.

And the fact that the litigation was over copyright is an insignificant detail. It could have been anything. Literally anything, like a murder investigation, for example. It only helps OpenAI here, because it's easy to say "nobody cares about copyright", and "nobody cares about murder" sounds less defendable.

Anyway, the issue here is not copyright, nor "AI", it's the venerated legal system, which very much by design allows for a single woman to decide on a whim, that a company with millions of users must start collecting user data, while users very much don't want that, and the company claims it doesn't want that too (mostly, because it knows how much users don't want that: otherwise it'd be happy to). Everything else is just accidental details, it really has nothing to do neither with copyright, nor with "AI".

Since 70% of comments under each of these posts are always "why not LaTeX?!", I want to start by reaffirming: yes, I do want a modern Markdown-based typesetting system. There even is a place of several of them. It would be absolutely nice to replace LaTeX, simply because it's old trash with remarkably inconvenient syntax, so, yes, a system with full control over markup is desirable. And if it necessarily increases verbosity, then there absolutely is a place for something "just a touch more powerful than markdown".

However, at a glance it doesn't strike me as what I was looking for. There aren't too many examples, but it seems like it leans heavier towards "just a touch more powerful than markdown" rather than to "replace LaTeX" (or Typst, for that matter). And for the first scenario to play out, it must be really seamless to use. This one doesn't seem to be.

1. Just to take care of the elephant in the room: JVM. Same as many others, I won't even bother installing it to try it out, so it doesn't help virality much.

2. I don't like the syntax. It needs to be as compatible with plain markdown as possible, and this one isn't quite. My main issue is with argument to a function being tabulated. It seems like it will lead to the whole document being tabulated. It is a natural for markdown-compatible add-ons to employ a code/monospace blocks, but ```plugin-name is a better way to do it, because it doesn't mean you have to reformat your whole document when you decide to step off from plain markdown.

3. Since a "better markdown" is something more suitable for something that starts as your personal notes (if you are specifically working on preparing a document for publishing, indeed you can just do it in LaTeX as well), I suppose it's only useful to an extent it is integrated into your notetaking app. Surely, there are still some people who do it in Emacs or Vim, but even such a retrograde as myself eventually moved on to Obsidian. I want ways to control my document structure better inside my notetaking app with a potential ability to publish. But as a standalone thing, I'm not sure why would I use it. At least Typst has a proprietary online editor. I suppose, that's also how nearly everyone uses it.

There hardly is such thing as "discounts" in airline pricing. I mean, formally, there is, there's a lot of them, but, well, it's complicated…

In all honesty all this thread is people complaining about something they don't have a clue about. Airline pricing is insanely complicated, and this is for a reason. Airlines are not a luxury business, they barely manage to survive. If not all this dynamic pricing, special contracts with agencies, etc, they'd have to charge so much for a seat that you wouldn't pay and all this travel industry you are accustomed to simply wouldn't exist. The whole business is built on making somebody who crucially needs to fly pay as much, as he can, and then to make price attractive enough for the rest of us so that you can sell the rest of the tickets, so that flight can make any profit. And in the end, margins are super thin in this business.

Also, your question implies that you imagine that there is some simple enough "true price for a seat", which is so far from the truth, you have no idea. If you actually look at the price breakdown for a given ticket, there are literally dozens of components in it. It's not unusual that so called "fare" of a ticket (which is, like, "just price") may be literally $1, and the rest of $300 is various taxes, surcharges and payoffs I won't even try to start to explain here.

I mean, really, people here truly have no idea what they are complaining about. Airline pricing is not a thing you should hate.