It's very impressive!
I'm mean, it's better than other <200M TTS models I encounter.
In English, it's perfect and it's so funny in others languages. It sounds exactly like someone who actually doesn't speak the language, but got it anyway.
I don't know why Fantine is just better than the others in others languages. Javer seems to be the worst.
Try Jean in Spanish « ¡Es lo suficientemente pequeño como para caber en tu bolsillo! » sound a lot like they don't understand the language.
Or Azelma in French « C'est suffisament petit pour tenir dans ta poche. » is very good.I mean half of the words are from a Québécois accent, half French one but hey, it's correct French.
> GitHub should charge every org $1 more per user per month
It's about org, not about every single person using Github.
The idea is basic and should have been written in the article.
When a contributor release FOSS, it's fair to compensate if you business rely on it.
A contributor wouldn't like a free for personal use either.
The ideal license is the Unreal one free for « Individuals and small businesses (with less than $1 million USD in annual gross revenue) »
> you're kind of dumb for giving it away and that's your fault.
It happens so many times and no just about software (but then it's not a million dollar company). It's not that you are dump, you done the right thing and some companies with money/power/opportunity to capitalize on it, did it and didn't compensate you fairly.
> When a contributor release FOSS, it's fair to compensate if you business rely on it.
Nope.
Put it in the license, sell the software, or work for free, but stop complaining about it.
It's nice if businesses who benefit from specific software packages want to pay or show support, but it's not nice to release something "for free" but then jump on a moral grandstand and demand everyone pay so you can feel good about your ideology at the expense of everyone else.
> The ideal license is the Unreal one free for « Individuals and small businesses (with less than $1 million USD in annual gross revenue)
You're not wrong, but I feel like a lot of people are hung up on the purism of the OSI definition, and a license that's not so blessed may prevent a project from gaining significant traction, if that is part of their goal.
I think it would be nice if there were a license that was more widely accepted that introduced a monetary component that could compensate the developer(s).
I can confirm.
I found multiple problems in the "official" cjxl encoder back in 2023 contrary to the webp2 (cwp2) implementation where I could not find any bug or error.
If the encoder have obvious problems it is not a big deal, but it doesn't bode well for the decoder.
Hmm, you are referring to rich numbers but pointing to normal numbers, so I will be the nerd who points out that every normal number is rich, but some rich numbers aren't normal.
Snapdragon is already canceled so I guess, they just don't care about this device.
It's Microsoft on ARM. Sad to say, but don't expect full support or quality update on this.
The Snapdragon Dev Kit is canceled. Snapdragon as a whole sure as hell isn't canceled, and Windows on Snapdragon isn't, either. There's loads of Windows laptops using Snapdragon with more continuing to release.
It's just the "dev kit". Snapdragon for the laptop form factor is alive and well. You don't need a devkit for a laptop running Windows and QCOM easily figured that out.
I wanted to order one of these and then Qualcomm cancelled it.
Then I knew Windows ARM probably wasn't going to make it. Why any technical person would want a PC( not including Macs)that explicitly can't run Linux I'll never know.
Technical person that knows UNIX since being introduced to it via Xenix in 1993, and has used plenty of UNIX flavours since then.
Some of us like the experience of Visual Studio, being able to do graphics development with modern graphics APIs that don't require a bazillion of code lines, with debuggers, not having to spend weekends trying to understand why yet again YouTube videos are not being hardware accelerated, scout for hardware that is supposed to work and then fails because the new firmaware update is no longer compatible,....
Your comment appears to address the question "why use Windows" (even though the answer doesn't really make sense to me), but that's not the question asked in GP. The question was "Why buy a Windows on ARM device"
PCs aren't vertically integrated from a single vendor, and thus it isn't as if Microsoft alone can drag a whole ecosystem into ARM, even if the emulation would work out great.
Windows NT was also multi-architecture, and eventually all variants died, because x86 was good enough, and when Itanium came to be, AMD got a workaround to keep x86 going forward.
Even gaming doesn't work that great on Windows ARM.
They have the Surface line and own tons of game studios.
Where are the Gamepass games with Arm ?
Microsoft if they wanted to fund it right could get popular 3rd party software ported.
In retrospect it was hopelessly naive, but I even emailed Qualcomm asking if I could have a dev kit in exchange for porting one of my hobbyist games. They basically said thank you for asking but we don't have a program for this.
Now hypothetically let's say there was a Qualcomm Snapdragon Linux laptop. I could just port the code myself for most applications I actually need
These devkits are old and have already been released to consumer laptops over a year ago. So if you want to you can pick up pretty much any CoPilot+ PC. I'm not sure what your problem here is though.
Tuxedo is a german company relabling Clevo Laptops so far, which work out-of-the-box pretty good (I might say perfect in some cases) on Linux. They have done ZILCH, NADA, absolute nothing for Linux, besides promoting it as a brand. So now they took a snapdragon laptop, installed linux and are disappointed by the performance....Great test, tremendous work! Asahi Linux showed if you put in the work you can have awesome performance.
Yes but having to reverse engineer an entire platform from scratch is a big ask, and even with asahi it's taken many years and isn't up to snuff. Not to say anything of the team, they're truly miracle workers considering what they've been given to work with.
But it's been the same story with ARM on windows now for at least a decade. The manufacturers just... do not give a single fuck. ARM is not comparable to x86 and will never be if ARM manufacturers continue to sabotage their own platform. It's not just Linux, either, these things are barely supported on Windows, run a fraction of the software, and don't run for very long. Ask anyone burned by ARM on windows attempts 1-100.
> if you put in the work you can have awesome performance.
Then why would I pay money for a Qualcomm device just for more suffering? Unless I personally like tinkering or I am contributing to an open source project specifically for this, there is no way I would purchase a Qualcomm PC.
The original comment was "explicitly can't run Linux" which is explicitly not true. Not "it's not fully baked" or "it's not good", but a categorically unambiguously false claim of "explicitly can't run Linux" as if it was somehow firmware banned from doing so.
If someone wants to provide a link to a Linux iso that works with the Snapdragon Plus laptops( these are cheaper, but the experimental Ubuntu ISO is only for the elites) I'll go buy a Snapdragon Plus laptop next month. This would be awesome if the support was there.
It is terrible in the opening and a little better in the middle game (if you didn't win already, I mean if you played 4-5 random moves).
I think you can improve it a lot by taking into account its own position. I mean, moving forward isn't very important want you are under the menace of a checkmate somewhere else.
I tried playing against it, I didn't have many expectations, but even though I blundered a bishop on move 3 due to a mouse-slip, I could still checkmate it in 6 moves. To me it seemed like it makes random moves.
Chose the most aggressive move (in term of pieces value and check-mate), if none is aggressive, it takes on of the move equally non aggressive.
Didn't remember the depth of the algorithm but it was very simple C code, could check quickly. It should be able to find a mate in 2 or 3 if it was in position of having one.
I didn't check the correctness of the algorithm, just the intention.
I thought it played worse than random moves and couldn't understand how it could beat anyone (no offence to OP).
But if you intentionally hang your pieces, it tends to take them. And it will try to promote pawns in the endgame. So it is possible for it to stumble upon a checkmate, though in my effort where I gave away all my pieces, it instead found the only move to stalemate once it had K+Q+R vs K.
It’s terrible everywhere tbh, it has zero awareness of how dangerous the f7 square is in the opening, you can pretty much just mate it by landing the queen on f7 with a piece covering it every time.
I applaud the effort but not sure I see the point, I had a more capable chess playing program than this on a zx speccy in the 80’s.
Bart was a flop.
Google search is losing market share to other LLM providers.
Gemini adoption is low, people around me prefer OpenAI because it is good enough and known.
But on the contrary, Nano Banana is very good, so I don't know.
And in the end, I'm pretty confident Google will be the AI race winner, because they got the engineers, they tech background and the money. Unless Google Adsense die, they can continue the race forever.
If Google is producing very good models and they aren’t gaining much traction, that seems like a pretty bad sign for them, right? If they were failing with bad models, the solution would be easy: math and engineer harder, make better models (I mean, this is obviously very hard but it is a clear path). Failing with good models is… confusing, it indicates there’s some unknown problem.
It’s irrelevant, Google needs to focus on performance enhancements that the enterprise market segment demands - who only operate in the air of objectivity.
If they can achieve that they will cut off a key source of blood supply to MSFT+OAI. There is not much money in the consumer market segment from subscribers and entering the ad-business is going to be a lot tougher than people think.
I'm talking about adoption around me. I didn't check the data.
Glad it works for them outside of people around me, didn't know that.
Is it relative adoption or absolute ? I mean, the people using Gemini are they new or coming from another provider, like OpenAI ? (said differently: Is Google eating OpenAI lunch or just reaching new customers ?)
In English, it's perfect and it's so funny in others languages. It sounds exactly like someone who actually doesn't speak the language, but got it anyway.
I don't know why Fantine is just better than the others in others languages. Javer seems to be the worst.
Try Jean in Spanish « ¡Es lo suficientemente pequeño como para caber en tu bolsillo! » sound a lot like they don't understand the language.
Or Azelma in French « C'est suffisament petit pour tenir dans ta poche. » is very good.I mean half of the words are from a Québécois accent, half French one but hey, it's correct French.
Però non capisce l'italiano.
reply