DiogenesKynikos

It sounds like you need to be less condescending and engage with my actual comment.

First, the "hype" is not externally imposed. It's the genuine reaction of hundreds of millions of people to a technology that would have been considered science fiction just four years ago. The fact that I can't be certain you whether you are an LLM or a human is incredible.

Second, the output of the LLM depends on both its weights and the context. If it has seen something in its context window, it now knows it for all intents and purposes. It learns, in other words.

Gain of function research is not anywhere near as dangerous as the public believes. In the US, it was very convenient to blame it for the pandemic, even though SARS-CoV-2 is of natural origin and almost certainly spilled over at the Huanan wet market in Wuhan. The threat of viruses comes almost exclusively from nature, which is constantly cooking up new viruses all by itself and exposing people all over the world to them. A few people doing tightly controlled research under high-biocontainment are a drop in the ocean. But their research is the main way we can prepare to deal with future pandemics, not to mention understanding the usual viruses that already afflict humanity.

This is when you realize that English is a Germanic language.

"Shprakh" is from the same root as "speak," so the literal, word-for-word translation is, "A speak is a dialect with an army and flotilla."

I bought a cheap electric toothbrush from a no-name Chinese company. The battery lasts 6 months on one charge with daily use. It cost about $20. I wouldn't be surprised if it's manufactured by the same factories that make Philips' toothbrushes.

And of course, it charges with USB-C.

Louden points out, for example, that Swedish and Norwegian are highly mutually intelligible, but neither is considered a dialect of the other, or of a parent language, primarily because each is associated with a separate nation-state.

This reminds me of the famous saying, "A language is a dialect with an army and navy."

It was also originally uttered in a German-adjacent language, Yiddish: "a shprakh iz a dyalekt mit an armey un flot."

I wonder whether locals in the German Palatinate region can still understand Pennsylvania Dutch, given that it supposedly originates from their dialect.

You're making a highly arbitrary distinction between learning and ... learning?

The LLM can immediately learn and start using new skills. To any lay person, that's what learning is.

The LLM is the model + context. The output depends on both.

You're making an artificial distinction. The LLM sees a new programming language for the first time, and can immediately code in it. That's learning by any reasonable definition.

If you go past the context window, it forgets, which is a limitation of current LLMs. But as long as it learned how to code in the new language within its context window, it has gained that new ability.

In theory, it can't. But in reality, it does in important cases.

Germany has implemented EU sanctions against a German journalist, which deprive him and his family of the ability to conduct basically any economic activity in Germany or even to leave the country. He is not allowed to work. No one is allowed to pay him money. He has to petition the government every time he wants to access even a small amount of the money in his own bank account. The same restrictions apply to his close family members, because they are suspected of helping him survive financially. He is barred from crossing any border in Europe, including to leave.

He has not been accused or convicted of anything in court. The only procedure that was required to hand down an economic death sentence was for the EU Commission to put his name on a list.

The German government claims this is all okay, because the journalist can proactively challenge his sanction listing in Brussels. It's a years-long process that will require paying lawyers - using money he is not allowed to access.

Needless to say, this is all highly unconstitutional. But the German government simply doesn't care. They just say they're implementing EU sanctions.

https://en.wikipedia.org/wiki/H%C3%BCseyin_Do%C4%9Fru

I can imagine an AI insulting humans in the same way:

"The underlying model is just a biological neutral network. It seems you carbonoids get upset when someone talks honestly about synapses and neuron firing."

Is your position that all coverage of Rust is gossip? That's just not a tenable position.

By the way, the article I linked to discusses the history of Rust in the Linux kernel, which is hardly gossip and illustrates why Rust is notable.

You've finally found a better source than YouTube for Odin. Whether it meets Wikipedia's standards will probably depend on whether it's self-published.

I gave you the first search result. There are probably tens of thousands of articles in newspapers and magazines about Rust, along with probably hundreds of published books. Unlike Odin, Rust is notable and heavily covered by secondary sources.

This took me approximately two seconds to find: https://arstechnica.com/gadgets/2025/02/linux-leaders-pave-a...

Publishing retractions is a sign of reliability. Wikipedia is actually more concerned about sources that don't publish retractions.

If Odin is so notable, why haven't any of the normal tech publications ever mentioned it? One company using it does not necessarily make it notable.

Remember, CNN and all the "trusted" news outlets also have YouTube accounts.

And those YouTube accounts are not considered reliable sources by Wikipedia. In fact, basically only CNN's published articles are considered reliable.

Shifting the problem from anonymous Wikipedia editors to newspapers and magazines is a big improvement.

Do you want Wikipedia editors to start citing YouTube videos?

I'll tell you what will happen if sourcing standards are loosened in that way. You'll get tons of editors arguing, "So and so said this at 21:45 in this podcast." And then someone else will say, "But he said this other thing that contradicts that at 22:30." And then a third editor will come along and say the person was just joking. The signal-to-noise in the sources will go way down, and you'll have editors acting like detectives and investigative journalists with a mass primary sources. It will be a mess, and it won't work.

Those "weird websites" have a reputation one can track. One can look at whether they have a history of correcting errors, citing sources correctly, etc.

That's why Wikipedia relies on secondary sources with a history of reliability. It's a simple rule that works for a website with tons of anonymous editors.

And guess what? There are tons of secondary sources that cover every major programming language. It looks like Odin is highly niche, which is why it has no secondary sourcing. Maybe it doesn't need a Wikipedia article.

Wikipedia has many policies. Verifiability is one of the most important ones: https://en.wikipedia.org/wiki/Wikipedia:Verifiability.

In this case, Wikipedia's policy is reasonable. I tried looking for any articles in reputable sources that discuss the Odin programming language. I couldn't find any. Not one article in any tech magazine.

I tried looking for reliable secondary sources that discuss the Odin programming language. I couldn't find any.

That's not Wikipedia's fault. It is a bit weird that people are insisting that Odin is such a big deal that it must have a Wikipedia page, but apparently no tech website of note has ever mentioned it.

It's also that letting editors argue based on their own original research would be a recipe for chaos.

"Use reliable secondary sources" is a simple rule that limits the scope for possible conflict and chaos. For an online encyclopedia that anyone can edit, that's very important.

Because plenty of secondary sources have covered it.

Wikipedia's philosophy is that it summarizes reliable secondary sources. You're asking for Wikipedia to change its fundamental philosophy so that editors go out and gather primary evidence themselves, debate it, and then deduce what's true by themselves. That would work if Wikipedia were edited by a select group of highly qualified people you can trust, but it's not. It's edited by literally anyone who signs up. So the rules are strict: find a secondary source, or else don't include the info.

Wikipedia's core policy is verifiability, not truth. If you put a factual statement into an article, you're required to have a written source that Wikipedia can cite. "Someone who says they're the CEO said so on the talk page" doesn't cut it. If the CEO really wants to get this info into Wikipedia, they can go give an interview to a reputable newspaper, news website, magazine, etc., and then Wikipedia can cite that.

Which brings me to the second policy: notability. If JangaFX is really notable enough to have a Wikipedia page, then surely there will be plenty of coverage of it in secondary sources, so Wikipedia won't have to rely on talk-page statements by random editors claiming to be the CEO.

Switzerland is half as dense as New Jersey.

It's around the same density as Delaware and Maryland. And it's full of giant mountains, which neither Delaware nor Maryland have.

Switzerland is small but mountainous.

And just for reference:

  * Delaware: 200 people / km^2
  * Switzerland: 230 people / km^2
  * Massachusetts: 350 people / km^2
  * New Jersey: 490 people / km^2

Switzerland.

Switzerland has some of the highest prices in the world for everything. This isn't unique to the Internet.

The framing here is sensationalist.

Going from near zero 40 years ago to being just 10 years behind the state of the art now is actually pretty rapid convergence. Calling that a "failure" is strange.

Beijing - Shanghai is about 1200 km. HSR has a 70% market share. The trip takes 4.5 hours.

The only real reason you would fly this route would be as a connecting flight. But if you just want to travel from Beijing to Shanghai, HSR is the obvious choice. It probably takes less time overall, and it's way more comfortable. You can even order takeout from restaurants along the way, delivered to your seat.

Note that most of those subsidies are things like sales-tax exemptions for EVs and support for charging infrastructure in China.

In other words, they're not subsidies for Chinese cars being exported abroad. They're not even directly paid to the manufacturers.

"Pump and dump" is when Trump talks up a stock he just bought and then sells it.

Because the journal is a stamp of approval, and scientists are measured almost exclusively through their publication record. Getting a paper accepted by Nature can make a career.

It's easy for anyone to publish papers online, but it's very difficult for a journal to build credibility and reputation. If you publish in some random journal no one has ever heard of, everyone will assume your paper couldn't get through peer review at a "real" journal.

That's why the established journals exist.

Nice, so every time I visit a restaurant and want to leave a review, I'll have to lawyer up. Sounds eminently reasonable.