Actually the fact the inference of a SOTA model is completely Nvidia-free is the biggest attack to Nvidia every carried so far. Even American frontier AI labs may start to buy Chinese hardware if they need to continue the AI race, they can't keep paying so much money for the GPUs, especially once Huawei training versions of their GPUs will ship.
By "completely Nvidia-free" do you mean Nvidia wasn't used for training nor inference? Because if it's only inference, we know that Opus already can run on TPUs. Not to mention Gemini.
Yep but they don't run on Chinese hardware that is going to be available to everybody and will cost a lot less than NVIDIA stuff. So now you have a full non-US pipeline for AI, and soon they'll have the training GPUs as well.
In the article it is mentioned but it is worth stressing that N-acetylcysteine is a trivially available antidote for paracetamol overdose (and you may have it at home without knowing, Fluimucil, Mucomyst, NAC or alike).
Also: in Europe everybody normally takes paracetamol and not FANS as a first reach to minimize adverse effects. So this article looks like very US centric. AFAIK liver failure because of paracetamol in Europe is very rare. So here there could be cultural issues at play (medical culture of what is prescribed, and the fact that Europeans in general take lower dosages of everything).
EDIT: trick, if you very rarely take paracetamol and other pain medications, the next time try to take just 250mg. It works for most people, no need to take 750 or even 1 gram of paracetamol. 500 works for almost everybody, 250 for many folks.
> Also: in Europe everybody normally takes paracetamol and not FANS as a first reach to minimize adverse effects. So this article looks like very US centric.
This is not my experience. After moving to Germany from the UK, I feel like people take and expect Ibuprofen far more often than Paracetamol. It seems like the first port of call for colds and general headaches, with Paracetamol being treated with some suspicion, despite it being far more effective in my experience for certain things (I've taken a lot of Ibuprofen and other NSAIDs in my time so am quite familiar with how they affect me).
Didn't expect that, my feeling was that the situation in most continental Europe was like in Italy and the UK, and even more in France! (I believe they have the record of use), in such places paracetamol is very strong.
1. Big frameworks were pushed from companies to the user base, most of the software stack is, instead, organically created by the programming community. So we have things like Angular and React that are big-company-design stuff that became normal programming. It's like if every site runs on Kubernetes.
2. There was, in big companies, an extreme desire to do two things: totally isolate frontend from backend, because the internal organization of big companies has such a split, and to make applications so standardized that hiring new people, firing old people, is something possible and easy. Fast, highly reactive applications written using semantical HTML and just the amount of required Javascript are, instead, general code, and each app is a little different, or even a lot different. So those two goals are not aligned to producing great web apps, but to internal requirements.
3. We later created a generation of programmers that can't even understand a single language very well in its internals, that is: Javascript, they often know the framework, not the language, nor even CSS well enough. So they are trapped into the ecosystem of knowing React or Angular or something like that. It was enough to have a high paying job. Now, those folks, the way the were trained and the way they work makes of them not really general purpose programmers like you could find 10 or 15 years ago. And now the industry is kinda trapped into this situation also because of people abilities.
Of all that, in theory, I don't care much, I can write my stuff as I want. But:
A) The web is terrible now, it is slow, eats a lot of memory, and is fragile. So this impacts me.
B) The big salary front-end folks culture is also made of a lot of people that don't realize that they move in a very limited field of what they believe programming is, and are arrogant enough that if you do something in vanilla Javascript, fast, well coded and that works well, you will be accused to do things in a odd / old way. Many of them basically actively resist to making the web more sane.
I believe LLMs are going to change this thing, as now there is less direct exposition to the specific code, and things can be more easily refactored in versions of the same app that are saner. We will see. But nothing in programming, in the latest decades, was at the same level of mess as of front-end development.
The irony is that front-end developers highly suffer from all that, for a number of reasons: they are forced to continue learning new ways to do the same button, form, pagination, and so forth. And, also, if they are smart they understand they don't really know what programming really is in most cases, and are not happy about it.
1. I think it's pretty clear that the frameworks sold themselves, because it's possible to see the DX benefits with anything beyond a hello world. You don't need Meta's scale. It came from the big companies because they needed it most, not because they are the only ones who benefit.
2. I agree that the frontend-backend split caused a lot of harm and it seems most places are moving on, but it
3. It's a fair assessment that many people don't understand JS well. But it's also a super quirky language with many features that probably should be considered "do-not-use". Things like messing with the prototype chain, purposefully using == instead of === etc.
CSS is even worse. The defaults are weird, which is why everyone uses some form of CSS reset, many things have surprising names and refactoring in a codebase with complex usage of selectors and cascading is a nightmare.
But I wouldn't call people trapped in frameworks. It's pretty easy to switch between React, Svelte, Angular and co.
A) The gone by days of Java Applet's and Flash weren't so snappy either. Our websites also do much more. A tab of Google docs isn't slower or use more memory than a Word instance.
B) No group is ever immune to arrogance. But sure, if you write vanilla JS instead of TS, I'd think that's an odd choice. I'm not saying you can't make good things with Vanilla JS, but I don't see how that makes the Web more "sane".
LLM: When it comes to coding tools LLMs are not so different from humans in what helps them. They benefit just as much from the cleaner control flow of reactive frameworks (compared to jQuery), colocation of JSX and tailwind and typechecking of TS as us humans.
Mess: Has the Web really been such a mess lately? I think most of the messiness is far behind us. JS+HTML5 winning out against Flash and Java, standardization across browsers, XML or HTML.
Was it worse than Microsoft reinventing their Windows UI framework every couple of years?
It is very hard to evaluate Apple as a single entity. Certain branches of the company did great in the latest 10 years: the chips division being the most obvious one but in general hardware did well. Software received a lot less love, and in general deteriorated, but the really terrible performance was the products division. Then there is the horror story of Apple software stores and services.
Apple is held to a different standard. Tahoe is bad because it's ugly and has inconsistent corner radii. Meanwhile Windows has multiple critical issues e.g. Windows update induced instability.
Apple's hardware and silicon divisions are considered good not because they're competitive, but because they're uncompetitive. They keep cranking out products that far surpass their competitors. Apple is reduced to announcing the fastest CPU core in the world in a press release spec bump, every year.
Even if Tahoe is a regression from the high bar Apple set with Sequoia, it's still the best OS for regular people (sorry Linux but you know it's true). Apple's software is bad because it's merely best-in-class and not light years ahead.
This is not in antithesis. My limited personal experience is that I wrote code under OSS licenses primarily because of my past communist believes and current left-wing and redistribution of wealth point of view. This is not to provide the simple equation of: communist China is not interested in money, but also is hard to believe that there is no cultural connection among those things. Single Chine persons want to win, but also they have a different POV on what the collective means, compared to US. Also there is the obvious fact that in this moment China is more interested in winning technologically in AI, more than economically, since, I believe, they more collectively realized before many others that LLMs are eventually commoditized in the current form, in the long run. One could assume that a breakthrough could give some lab a decisive advantage, but so far we assisted to a different reality: it looks like AI is not architecture-bound (like LeCun and others want us to believe, but so far they mis-interpreted LLMs at every step) but GPU bound, and the data-boundness is both a common ground for all, and surpassable via RL in many domains. So, if this is true, it is not trivial for any single lab to do so much better. And indeed as far as we observed right now folks with enough engineers, GPUs, money, can ship frontier models, and in China even labs with a lot less GPUs can still do it at a SOTA level. For me, Italian, this is also a protective layer. After Trump the US looks like a very unstable partner from which to relay in an exclusive way for a decisive technology, and given that Europe is slow to put the money in this technology to have frontier things at home, China is a huge and shiny plan B for us.
The strings attached by the US to deep partnerships are things like trade/commerce, militarily mutual advantages (bases on euro soil from which we will help protect you), not to mention the close cultural and ancestral ties we share.
The strings attached by the Chinese govt to deep partnerships are not so benign.
Unfortunately the generation of the English audio track is work in progress and takes a few hours, but the subtitles can already be translated from Italian to English.
TLDR: It works well for the use case I tested it against. Will do more testing in the future.
Similarly most leds are photo diodes, electric motors can be used to generate current, Peltier cells can be used to generate current, and so so forth. Many of such physical processes are invertible.
Yes. Mostly the same basic components--Just optimized for which direction you are going in. I recall using an single in ear speaker as a microphone as one of the experiments in my Radio Shack 101 eperiments electronics kit.
reply