free_bip

It's funny you say that, because energy actually isn't conserved in general.

One somewhat trivial example is that light loses energy due to redshift since photon energy is proportional to frequency.

I'm going to steelman your argument and assume a few things:

- You are a real person who genuinely wants to make a difference in the field of physics

- The proposed mathematics isn't technobabble and actually does generate the constants you say it does

- The contributions are overwhelmingly yours, not overwhelmingly AI-generated

Even under these generous assumptions, this just isn't how citizen science is done. Your first step isn't publishing a 2-page paper with no references and basically no details on what you're doing or why. Your first step is taking your results to the nearest PHd theoretical physicist you can find, and getting them to take you seriously so you can find out if your work actually has merit.

I'm having trouble understanding the use case for this. I can't remember ever using a QR code as a way to verify the authenticity of something. It's always been something like "here's a link to my website" or "here's a movie ticket that's scannable in a physical theater". So what is this solving?

Looking at the paper, the only referenced source is grok. So yeah, ai slop.

I guess I'm worse than a fool then, because I thought it was totally impossible 10 years ago.

Their technique does not claim to compete with gradient descent - it's competition for techniques like Proximal Policy Optimization, so it's more suited for things like creating a reasoning model out of an existing pre-trained model.

Person or painting is not relevant in this argument by analogy. It's the commonality of "more than meets the eye", or "don't judge a book by its cover", or "appearances can be deceiving" between the two situations. There's dozens of idioms for this situation because it's so common.

I could just as easily make a painting analogy. You go to an art gallery and a piece immediately catches your eye. The composition, colors, you even think you can see a rosy meaning behind it all.

Then the curator comes by and explains that it was painted by David Duke, grand wizard of the KKK, and that the piece represents his deeply racist hatred of black people and his desire to murder innocents.

What you feel after having that explained is not "post hoc rationalization" - you're just recontextualizing your previous feelings towards the piece now that you have new information. That is the argument.

Let's say I introduce you to a friend of mine. You chat for a little while, and everything seems to be going well, you like them. Then, I tell you that my friend loves kicking puppies and infants. Really hard too, like he's kicking a field goal.

All of a sudden, you start acting like he's a "bad guy," who should "burn in hell." You liked him just fine before I gave you that info, so clearly that's just a post hoc rationalization that cannot withstand the simple fact that he was liked as a person before his interests were known.

It's a huge piece for sure, but not the only one. For example, Firefox and Windows both don't support it out of the box currently. Firefox requires nightly or an extension, and on Windows you need to download support from the Microsoft store.

Do you really need 17 helicopters and 90 employees for the occasional car chase? This feels wildly over the top. They could do that just fine with one third the resources.

Yeah? What about them?

You are not immune to propaganda.

Maybe if you're using their site directly, but what about the open models?

If your argument is that all of these things somehow combine to make the specific case I mentioned in my original comment legal (which was "stealing the work of every single artist, living and dead, for the sole purpose of making a profit", and I'll add replacing artists in the process to that), then I'm not seeing it.

You also seem to be talking about AI training more generally and not the specific case I singled out, which is important because this isn't a case of simply training a model on content obtained with consent - the material OpenAI and Stable Diffusion gathered was very explicitly without consent, and may have been done through outright piracy! (This came out in a case against Meta somewhat recently, but the exact origins of other company's datasets remain a mystery.)

Now I explained in another comment why I think current copyright laws should be able to clearly rule this specific case as copyright infringement, but I'm not arrogant enough to think I know better than copyright attorneys. If they say it falls under fair use, I'm going to trust them. I'm also going to say that the law needs to be updated because of it, and that brings us full circle to why I disagree with the article in the first place.

I think your analogy is a massive stretch. `wc` is neither generative nor capable of having market effect.

Your second construction is generative, but likely worse than a Markov chain model, which also did not have any market effect.

We're talking about the models that have convinced every VC it can make a trillion dollars from replacing millions of creative jobs.

I'm not a copyright attorney in any country, so the answer (assuming you're asking me personally) is "I don't know and it probably depends heavily on the specific facts of the case."

If you're asking for my personal opinion, I can weigh in on my personal take for some fair use factors.

- Research into generative art models (the kind which is done by e.g. OpenAI, Stable Diffusion) is only possible due to funding. That funding mainly comes from VC firms who are looking to get ROI by replacing artists with AI[0], and then debt financing from major banks on top of that. This drives both the market effect factor and the purpose/character of use factor, and not in their favor. If the research has limited market impact and is not done for the express purpose of replacing artists, then I think it would likely be fair use (an example could be background removal/replacement).

- I don't know if there are any legal implications of a large vs. small corporation using a product of copyright infringement to produce profit. Maybe it violates some other law, maybe it doesn't. All I know is that the end product of a GenAI model is not copyrightable, which to my understanding means their profit potential is limited as literally anyone else can use it for free.

[0]: https://harlem.capital/generative-ai-the-vc-landscape/

One of the few times I vehemently disagree with the EFF.

The problem is this article seems to make absolutely no effort to differentiate legitimate uses of GenAI (things like scientific and medical research) from the completely illegitimate uses of GenAI (things like stealing the work of every single artist, living and dead, for the sole purpose of making a profit)

One of those is fair use. The other is clearly not.

As much as I'd like to, the vast majority of android phones are incapable of installing grapheneOS.

Rosetta is a software translation layer, not a hardware translation layer. It doesn't take any die space.

What an insane statement given it has literally happened AND been a massive news story for quite a while now. For anyone still unaware of the massive extent of constitutional rights violations that dyauspitr mentioned, Legal Eagle has a several-part video series that I highly recommend.

But they literally are... Instead of hiring a Japanese person, they're hiring someone from the Philippines.

Because the experts have no say in policy. The only people who have a say are the people bribing (sorry I mean "lobbying") Congress. And even they have very little say because Congress is currently on a hot streak of doing absolutely nothing.

So the Ballmer peak is real?

It makes it to market because the US Congress is effectively nonexistent. They can't pass a bill, even a simple budget bill. The government is literally shut down.

I presumed they would leak apple juice.

webp

Just because there's only one deallocation doesn't mean it's run only once. It would likely be run once every time the thread it belongs to is deallocated, like when it's finished processing a request.

I disagree with the reasoning "Because there's too much important software that runs only on windows."

My disagreement isn't because wine or proton exist, it's because most people only use a web browser. They check their email, watch tiktok and netflix, and write documents. 90% of people would have all their computing needs met by a basic chromebook.

The funny thing is you did the exact same thing in this comment as the last one! No arguments to be seen, just "I did all this stuff." Maybe we should call this sunken cost rather than muddying the waters?

How exactly is that relevant? I doubt most people who use Amazon admire Andy Jassy.