applfanboysbgon

And why, pray tell, would a cabinet member of the Trump administration be involving the US government in enforcing a private ToS?

IMO, it's just the notation. Something I've actually found ChatGPT useful for is to create mathematics lessons for me in the form of computer programs. When broken down into a series of readable almost-plain-English steps, it's so much easier to understand. And it's easy to tinker with programs and get a hands-on feel for things quickly.

I'm sure having a compact notation is absolutely invaluable for people who dedicate their lives to maths, but for someone with just a passing interest, I find it more obscuring than helpful. I feel the same way about music notation.

For now. We've seen this pattern play out literally a hundred times in tech and you're incredibly naive if you think this will last forever. And what is your point? It should be okay for US consumers if the US government illegalizes accessing open-weight models within their borders because they're currently getting a subsidized token rate?

This is a misrepresentation of Singapore's policy. Nobody is being "shoved" anywhere, it is voluntary. And it is possible to replicate elsewhere. Taiwan reached 88% at one point; the Soviet Union was also known to have virtually eliminated homelessness, although in that case your description would have been accurate. Still, a home is better than no home. Government policy absolutely should be making a baseline of cheap housing available for everyone, even if it's not a 5 bedroom McMansion on an acre of land.

It's absolutely hilarious that vibe coders see themselves as some sort of protected class who are being discriminated against and think they can use the law to force an unrelated third-party to host their files for free. AI psychosis is a trip.

my point is that you can't boil happiness down to a single causal factor

This is not a claim I ever made. The claim I made was that homeownership was of significantly higher importance than affordability of cheap consumer goods. Obviously there are other factors.

though the effort to validate is likely higher than just accepting it.

Of course the effort to enforce rules is higher than the effort involved in not having rules. What is your point?

AGENTS.MD and CLAUDE.MD also provide easy detection avenues. Trivial measures will probably catch 70-80% or more. You don't need a 100% enforcement rate for a rule to exist, indeed there are basically no rules anywhere in the world that achieve that.

I suppose, if your completely insane fantasy ever becomes reality, they could simply reverse the policy. Codeberg currently lives in the real world, not your fantasy world, so it makes sense to determine policy based on the state of the real world.

If worse comes to worst, you can move another hour or two away from the big city.

It is rarely the case that someone who can't afford housing in their current location can afford to move away from their support networks to a place with no jobs. Places with significantly cheaper rent almost invariably give you significantly less income, if you can even get a job without connections (which are important in general, but especially important in rural areas where everyone knows each other). That can work out for you if you have substantial savings or a comfy remote tech job, but if you have those you probably don't need to move in the first place.

Your other points are valid ways to technically scale housing down but, I think, bring with them a significant decrease in quality of life that far outweighs what is gained by being able to afford 500 big macs and 10 pairs of shoes.

Singapore is surveyed as either the happiest or second-happiest country in Asia (IIRC), with at one point a >95% homeownership rate due to government policy (although I believe that's now closer to 90%, in part due to foreign immigration?). It seems self-evident that having a home is somewhere on the order of 1000000x more important than being able to buy lots of cheap junk that barely matters.

most of reality is pretty dull looking!

This is extremely untrue. It may be true that you spend most of your time in a dull-looking office or apartment, but most of reality -- at least here on Earth -- is quite visually interesting, and games are usually placed in visually interesting environments rather than set entirely within an office or apartment.

Beyond that, I would argue that the entire point of realism is not looking good (although it can look good), but rather immersion. The intent is to transport you to another world and make you feel like you're actually inside the scene. If your goal isn't to do that, then there's no need to bother with realism at all, and you can instead opt for highly stylized graphics that are intentionally designed to maximize aesthetically pleasing appearance and gameplay readability.

ShinyHunters has been "in business" since 2019 and it is their reputation that resulted in eg. Canvas paying their ransom this year. Without that reputation, it is unlikely a large-scale ransom would have been paid, because the reputation is what gives them credibility that paying the ransom will actually result in the promise being upheld.

One time I tried to write Doom for a platform, from scratch. I got the 3D BSP renderer working and could move around a space with walls. I was looking for reference for how to implement the rest of the game and didn't find it. I didn't realize you're supposed to port Doom to a platform by starting from the original Doom code and not by writing it from scratch, because the only detailed enough reference is the code itself.

This is the greatest misunderstanding of a programmer's job I've ever seen (by someone who apparently programs). Of course you don't write a game from scratch every time you port it to a new platform. You aren't even really supposed to be "referencing" the game code. The game code is built on a series of abstractions over the hardware. You swap in a set of abstractions for different hardware. If the game is properly encapsulated, you may not even need to touch what would be considered the "game" code rather than the engine code, or only minimally to update API boundaries.

eg.

Layer 0: Hardware

Layer 1: OS

Layer 2: DirectX

Layer 3: Engine code, has functions like Draw3D(params) which call DirectX functions

Layer 4: Game code, which calls Draw3D(params)

If you swap in OpenGL for DirectX, you simply change engine code so that Draw3D calls OpenGL instead of DirectX as appropriate, and the game code is still just happily plopping along with the exact same Draw3D calls ignorant of whether those are being served with DirectX or OpenGL. This applies all the same even if Layer 2 is a software renderer, as Doom's was. Ditto for OS layers or anything else.

If you want to reverse engineer it from scratch for learning, have at it, but why in God's name would you think that's what porting entails? And even after you understand porting is not about writing from scratch, you should understand that this has no relevance to your overall point. I'd also note that it is possible to reverse engineer a game from scratch without ever referencing the original source code -- that's called a decompilation project, it's intensive and time-consuming work, but it can be done.

Okay...? Do you believe you're entitled to everyone's code for free?

Saying “solve this problem” doesn’t get good results most of the time with humans either

Sure. That is not even remotely the point I was getting at. Already we see the thread filling up with comments about how human skills are irrelevant, using a mathematics PhD applying his expert skills in a way that the people who are saying that could never have done to justify their inane conclusion.

Two points:

- Hasn't been peer reviewed yet, so take with a grain of salt. This applies to all claimed proofs, not just AI-generated ones. Even humans hallucinate proofs too!

- The prompt is on page 27 here[1]. It is ten pages of advanced mathematics priming the model in the right direction, apparently informed by a year of prior research. That doesn't invalidate the result if it is genuine, but it is worth noting that this wasn't a matter of "ChatGPT, solve this unsolved problem. Make no mistakes." and required substantial domain expertise and human research beforehand.

[1]https://arxiv.org/pdf/2607.13335

In the end, IT DEPENDS. But there's apparently no room for such nuance at Recurse...

"One thing that often surprises people about the social rules is that we expect people to break them from time to time."

Sounds like there's plenty of room for nuance. They're guidelines. That you spent several paragraphs considering and articulating the difference between a positive social interaction and a negative one means the guideline is probably a good one, even if they didn't bother to write several paragraphs of nuance for each point themselves (which is perfectly reasonable, because brevity has its own value).

---

Rate-limit edit:

That they haven't done so in 15+ years suggests that they think the rule is well-written.

Or it suggests they know more about what social problems actually arise in their community. Your 'vastly better' suggestion of "Don't be a dick" is so vague as to be a completely meaningless guideline, and someone reading that could easily end up doing the condescending surprise act without thinking they're a dick. The purpose of the rule is to call attention to a specific behavior they likely see often to make you think twice about it.

How is it extraterritorial overreach when you acknowledge in the same comment that you can ignore it because you aren't in the EU? Unlike the site that is the subject of the thread, which has actually been subjected to extraterritorial overreach. This criticism and false equivalency is ridiculous.

Honestly, I wish more of the discussion had been about the technical side of the project.

Then you would do well to spend a few hours during your three year project to write up a technical description that indicates you actually understand what you've published. Saying you don't speak English natively is a cop-out. LLMs are subpar at translation but if prompted to translate correct technical writing, the output would not have been what's on your front page. You can't possibly present poor technical writing and then expect people to give you more than what you gave them. And yet, I think people did spend more time writing about the technical writing than you did in prompting.

At some point I realized that arguing about whether a project was "made by AI" or not is a bottomless pit

Using AI to polish a website doesn’t mean the compiler itself was generated by AI. I think that’s an important distinction.

I notice you don't actually say this isn't AI-generated. You try to heavily imply it isn't with selective wording, yet stop short of denying that it is. And with a fully AI web design, fully AI descriptions, incorrect technical writing, no source code - why would anybody believe otherwise?

Windows also has shortcuts to retrieve off-screen windows, but it's more difficult to remember so I have to google it every time this situation comes up.

Codex CLI is written in Rust and significantly more efficient than Claude Code. I had no trouble running it on a 2012 laptop. VSCode/standard browsers were the main problem software, as one would expect.

There is no difference in principle. That is equally unacceptable.

There is a difference in kind, because it becomes impossible for the global internet to exist if thousands of local jurisdictions are being given their way, with conflicting local legislation resulting in global takedown when it is impossible to comply with two different jurisdictions. So this is noteworthy as an escalation of an already existing problem into an even worse direction.

---

Rate-limit edit:

Which is why the Internet hasn't been global in a long time, and looks pretty different in China vs EU vs Russia.

China and Russia aren't part of the global internet because they have national firewalls and segregated themselves. The EU very much is, and with limited exceptions the internet doesn't look much different from the US, the EU, Japan, Canada, Australia, Mexico, Thailand, Brazil, or South Africa. It seems absurd to suggest that the internet isn't global when I'm in all likelihood talking to you from the opposite side of the world and this is the norm. And what is the point you're making? That we should embrace the China/Russia model and give not only every country but also every state/province/city its own Great Firewall?

GDPR Article 17 expressly requires the removal of things from the global internet

...as part of compliance with GDPR, if you choose to be compliant. Please name one instance of the EU suing and successfully removing an American website from the internet under this article, or any part of the GDPR? Considering we're talking about an actual case of the US seizing the domain of a European website, whataboutting a hypothetical with the GDPR which has never done the reverse despite being in force for 10 years is incredibly disingenuous.

---

Rate-limit edit:

Are you saying that you don’t think that the GDPR text is written to apply outside of the EU, or that it does say that but it’s not relevant because it’s not viable for anybody to enforce that?

The GDPR is European legislation, written for the territory the EU has legal jurisdiction over. Why would anybody think it's meant to apply outside of the EU? Plenty of businesses choose to operate by two sets of privacy policies, one where they continue fucking over their American users and one where they adhere to the GDPR for European users, and that is perfectly acceptable. There is no "think" about it, the legislation obviously does not apply outside the EU, nor is it intended to.

The GDPR doesn't try to remove anything from the global internet. You're free to not serve your site in the EU. Texas is free to block sites in Texas. But Texas trying to remove the ability of Europeans to access European websites is a completely different matter.

It is jarring to read Claude talking like it uses credit cards or feels guilt over using reward points.

FWIW, the site isn't blocked globally. They just moved to a new domain.

I do generally agree that local governments trying to forcefully exert their influence beyond their jurisdiction is deeply problematic. It wouldn't even be possible to host a website on the internet if this becomes normalized, due to being held to thousands of contradicting standards. At most Texas should have the authority to tell Texas ISPs to block traffic.

The JetBrains team didn't feel like prompting for a Claude UI to go with their project they wrote by hand 8 years ago, I suppose. Are you really that impressed with stock LLM-generated UI?

Note the JetBrains playground includes features that are actually useful for a code playground, letting you configure compiler version and flags, which is likely more relevant than a faux-IDE UI for the use cases people use playgrounds for. It also supports importing packages, while this doesn't. This doesn't seem like a human even attempted to use it for anything, otherwise they would have realised how lacking it is. Unfortunately it seems like people who prompt slop can't be bothered to dogfood it... or maybe pigfood it?

"From scratch in C" means literally nothing here. LLMs can output C too. It may or may not be safe, efficient C, but writing C that compiles is trivial. Prompting "write it in C" poses no greater difficulty than any other language.

Chances are they aren't a technical person and didn't build it, so wouldn't have anything interesting to say anyways. Web compilers are dime-a-dozen and LLMs can easily produce them with no active guidance. Very much in the training data[1]. This looks like just another person posting something they spent all of two minutes prompting.

[1] https://github.com/JetBrains/kotlin-playground

Well, usually advertising themselves. Spambots used to be really straightforward. "Come buy our penis enlargement pills at x.com!".

I guess this is also meant to be an advertisement but it's really roundabout. And while this case is maybe an ad, a lot of these LLM spam articles are on empty blogs not associated with a product.