Literally nothing. A corporation, especially a publicly traded one like that, can’t do much but maximize (ideally long-term, but usually short-term) shareholder returns.

The Activision-Microsoft merger is a good recent example of this. During the anti trust trial, the CEO of Activision literally came out and said that he believes it’s a bad idea that will be bad for the industry and bad for the company in the long term, using the impact of consolidation in Hollywood as an example, but he has to side with the board. He’s basically legally obligated to.

I’m not saying it’s unjust or a bad system (and I’m definitely not trying to paint Bobby Kotick as a good guy), I just want to point out that corporations are very simple in their purpose, and nobody should be expecting anything more from them. If you’re disappointed that Google made this 180, that’s on you for falling in love with a corporation. They’re useful tools for producing goods and services, but terrible as a political tool for democracy.

But for some reason, it became popular to fetishize tech companies, and that spawned megalomaniacs like Elon, Zuckerberg, Horowitz, Thiel, etc who feel like they should be the supreme rulers of our civilization.

To my knowledge Musk is gambling with his own money, not hedge fund capital or something.

Well, he had to get financing from multiple banks to purchase Twitter

Late as fuck but I can’t allow a Lord of the Rings reference to fall through the cracks like this.

https://lotr.fandom.com/wiki/Gríma

Calculators made mental math obsolete. GPS apps made people forget how to navigate on their own.

Maybe those are good innovations or not. Arguments can be made both ways, I guess.

But if AI causes critical thinking skills to atrophy, I think it’s hard to argue that that’s a good thing for humanity. Maybe the end game is that AI achieves sentience and takes over the world, but is benevolent, and takes care of us like beloved pets (humans are AI’s best friend). Is that good? Idk

Or maybe this isn’t a real issue and the study is flawed, or more realistically, my interpretation of the study is wrong because I only read the headline of this article and not the study itself?

Who knows?

This doesn’t account for blinking.

If your friend blinks, they won’t see the light, and thus would be unable to verify whether the method works or not.

But how does he know when to open his eyes? He can’t keep them open forever. Say you flash the light once, and that’s his signal to keep his eyes open. Okay, but how long do you wait before starting the experiment? If you do it immediately, he may not have enough time to react. If you wait too long, his eyes will dry out and he’ll blink.

This is just not going to work. There are too many dependent variables.

There’s so much misinfo spreading about this, and while I don’t blame you for buying it, I do blame you for spreading it. “It sounds legit” is not how you should decide to trust what you read. Many people think the earth is flat because the conspiracy theories sound legit to them.

DeepSeek probably did lie about a lot of things, but their results are not disputed. R1 is competitive with leading models, it’s smaller, and it’s cheaper. The good results are definitely not from “sheer chip volume and energy used”, and American AI companies could have saved a lot of money if they had used those same techniques.

The model weights and research paper are

I think you’re conflating “open source” with “free”

What does it even mean for a research paper to be open source? That they release a docx instead of a pdf, so people can modify the formatting? Lol

The model weights were released for free, but you don’t have access to their source, so you can’t recreate them yourself. Like Microsoft Paint isn’t open source just because they release the machine instructions for free. Model weights are the AI equivalent of an exe file. To extend that analogy, quants, LORAs, etc are like community-made mods.

To be open source, they would have to release the training data and the code used to train it. They won’t do that because they don’t want competition. They just want to do the facebook llama thing, where they hope someone uses it to build the next big thing, so that facebook can copy them and destroy them with a much better model that they didn’t release, force them to sell, or kill them with the license.

Is it just me or is that title worded as confusingly as possible?

I’m seeing people say that the broadcaster (Fox Sports, of course) injected cheers into the broadcast for Trump, and boos for Taylor Swift. I don’t want to spread misinfo though so does anyone know if it’s true, or if there’s a way to validate it? (Eg by analyzing the audio)

EA should split itself up or sell off some franchises. The current situation makes nobody happy: investors don’t like the profitability of single player games, players don’t like the live serviceification of single player games, and I’m sure devs don’t like that they can’t work on projects that likely inspired them to become game devs in the first place.

The Sims franchise could support a medium sized studio on its own.

I played the Sims 1&2 as a kid, and love the fuck out of them. As an adult with disposable income, I would have gladly dropped even $100 on a proper modernized rerelease of these games I love. Instead, I saved my money and downloaded them for free. Because why the hell wouldn’t I? The pirate versions are literally better. EA is squandering the potential of this and many other IPs

BaCk iN mY DaY wE SaVeD To FlOpPy DiSkS!!

etc

96 GB+ of RAM is relatively easy, but for LLM inference you want VRAM. You can achieve that on a consumer PC by using multiple GPUs, although performance will not be as good as having a single GPU with 96GB of VRAM. Swapping out to RAM during inference slows it down a lot.

On archs with unified memory (like Apple’s latest machines), the CPU and GPU share memory, so you could actually find a system with very high memory directly accessible to the GPU. Mac Pros can be configured with up to 192GB of memory, although I doubt it’d be worth it as the GPU probably isn’t powerful enough.

Also, the 83GB number I gave was with a hypothetical 1 bit quantization of Deepseek R1, which (if it’s even possible) would probably be really shitty, maybe even shittier than Llama 7B.

but how can one enter TB zone?

Data centers use NVLink to connect multiple Nvidia GPUs. Idk what the limits are, but you use it to combine multiple GPUs to pool resources much more efficiently and at a much larger scale than would be possible on consumer hardware. A single Nvidia H200 GPU has 141 GB of VRAM, so you could link them up to build some monster data centers.

Nivida also sells prebuilt machines like the HGX B200 which can have 1.4TB of memory in a single system. That’s less than the 2.6TB for unquantized deepseek, but for inference only applications, you could definitely quantize it enough to fit within that limit with little to no quality loss… so if you’re really interested and really rich, you could probably buy one of those for your home lab.

If all you care about is response times, you can easily do that by just using a smaller model. The quality of responses will be poor though, and it’s not feasible to self host a model like chatgpt on consumer hardware.

For some quick math, a small Llama model is 7 billion parameters. Unquantized that’s 4 bytes per parameter (32 bit floats), meaning it requires 28 billion bytes (28 gb) of memory. You can get that to fit in less memory with quantization, basically reducing quality for lower memory usage (use less than 32 bits per param, reducing both precision and memory usage)

Inference performance will still vary a lot depending on your hardware, even if you manage to fit it all in VRAM. A 5090 will be faster than an iPhone, obviously.

… But with a model competitive with ChatGPT, like Deepseek R1 we’re talking about 671 billion parameters. Even if you quantize down to a useless 1 bit per param, that’d be over 83gb of memory just to fit the model in memory (unquantized it’s ~2.6TB). Running inference over that many parameters would require serious compute too, much more than a 5090 could handle. This gets into specialized high end architectures to achieve that performance, and it’s not something a typical prosumer would be able to build (or afford).

So the TL; DR is no

I switched to and exclusively used vim for about a year. I switched back to Sublime one day, and found I was like 10x more productive and comfortable.

Just use the editor you like. There’s no right or wrong answer!

… And btw, Sublime 4 has improved LSP support. Just install the base LSP plugin + plugins for the languages you want. Some even give the option to install the LSP server automatically if if’s not detected.

And on top of it all, he would just get a presidential pardon

Fuck yeah.

For anyone out of the loop, look into Freetube and Grayjay. There are other apps that do the same thing too, but those are good to start with.

There’s some politics involved. Basically, everyone is rallying behind JPEGXL instead of WebP, but Google refuses to support JPEGXL in Chrome. The reasoning they gave is weak, so it’s assumed that they’re just trying to force the format they invented on everyone because they can.

IIRC, performance of the two formats is similar.

Lazy motherfuckers on this site can’t even use proper grammar when being a snarky asshole. That shit you wrote is barely coherent.

logic error on line 2: Beer == Germans

Beer does not equate to Germans, rather Germans equate to Beer. If we fix that error, then it doesn’t fit the original pattern:

• Germans == Beer
• Germans == Fascists

That would only work if Beer == Fascists, which of course is not true.

Also, wrong does not equal stupid, rather stupid equals wrong. Which is to say, you comment is wrong, but not necessarily stupid.

Using a phone that long is risky due to the lack of security updates, especially if you’re using it for work. People not using phones longer is a problem, but the bigger issue is manufacturers killing support so quickly to force people into upgrading.

I recently upgraded after 5 years on an iPhone because it reached the end of its support cycle. I considered another iPhone because 5 years of support is great, but really didn’t feel like paying another $1000+ for what is essentially the same phone I was already using, just with a different body. So I went with a used Pixel 7 on ebay and installed GrapheneOS on it, and I’m very happy with it. I’m getting the same 5 years of support, a more secure OS, and I’m recycling at the same time!