I would assume these are entirely net new use cases. I also don’t know why this wouldn’t be developed in localized cloud regions. Aws, for instance, running in India is locally compliant with regulations. It’s cheaper for aws to comply with Indian regulations than it is for a random company, or worse a government, to build aws from whole cloth in India (nothing here is specific to India, just used as an example).
Before anyone steps in in favor of on prem vs aws, I’ve worked at cloud providers, building megacorp on prem, and migrating said on prem to the cloud, and building native cloud companies. Execution risks and the enormous high energy state of building full service on prem solutions make it much more reasonable to use cloud providers for stuff like this, and the economics of capital investment, the accounting of things, etc, makes it unreasonable for governments to NOT use cloud providers.
That’s only if there is an overlap between the customers who want to use cloud and the ones who want to run things on premise.
I would assume that those who want to run things on prem aren’t currently cloud customers for other services. And those who are already heavily invested in a cloud provider will look to the same provider for AI compute.
Go and meditate. You will never find truth in thoughts/words/code/beliefs. It can be in trillions of possible combinations. You can keep churning it out for infinity and not reach anywhere.
If they make the tech accessible for millions of people and improve reality, how is that a bad thing? I am confused. Unless you are too much emotionally caught up in the word "steal" and bunch of sentences.
Patents are a pragmatic contract between inventors and society:
- In order to be granted a patent, inventors must publish detailed descriptions of their inventions, including step-by-step instructions on how to re-create them. The inventions become public knowledge.
- In exchange for publishing the details of their inventions, society grants a temporary exclusive right to market the technology.
Without this system, the only way to safeguard a technological advantage would be to keep it secret. Inventors would be incentivized to maintain maximum secrecy for all new inventions. That would be bad for society, and inconvenient for inventors.
The patent system is an acknowledgement of this reality. Even though it seems a bit ridiculous, we allow people to claim temporary "ownership" of their new ideas, in order to incentivize the publishing of research, which eventually enters the public domain.
Clearly, there's some room for argument on the details of this specific case, but assuming the patent is legitimate, and that Apple infringed on it, then it would be bad for the long-term progress of science for us to let them get away with it.
> Without this system, the only way to safeguard a technological advantage would be to keep it secret. Inventors would be incentivized to maintain maximum secrecy for all new inventions. That would be bad for society, and inconvenient for inventors.
And in today's world where reversing a product is pretty much a guarantee if the product is something anyone wants, the secret won't be kept for too long. (shhh, everyone's secret sauce is pretty much thousand island) Once the secret is out, anyone that wants to will copy it and now there's no protection.
Ya, clearly patents are a good deal for inventors. I just think a lot of people fail to realize that it's also a good deal for society.
Without patents, who knows how many ideas would be stuck in the heads of individual people, reluctant not only to try to bring them to market, but to even discuss them with anybody. Why would anybody want to become an inventor, if there were so little upside?
to be fair, it's harder and harder to come up with an idea first. there's a lot of patents that have to be researched to see if your idea is first or just a first^Nth person to have the idea. that's a definite specialty skill and even then they're not perfect
The problem is many patents are incomplete, they don't actually contain step by step recreation instructions and it's especially problematic in the sciences fields.
The other issue is there are patents on ideas that aren't novel. There was some tightening of the rules such as the Alice software ruling but still lots of holes.
Case in point personally, our competitor patented the "idea" of putting a flexible PCB in a type of product. That's how bullshit it was and that was granted in 2020. Luckily we have prior art of a flexible PCB in a product of ours from 17years ago if they decide to ever start shit.
I bet there are lots of small time inventors that lack the capital to setup a worldwide manufacturing and distribution infrastructure right off the bat that may disagree with you.
Apart from those words, is there anything else you think is of concern compared to the progress made by humanity by making the tech available to millions of people?
News media like NYT, Fox etc are tools for high scale brainwashing public by the elite. This is why you see all the News papers have some political ideology. If they were reporting on truth and not opinions they won't have the need for leaning. Also you never see the journalists reporting against their own publication.
Humanity is better off without these mass brainwashing systems.
Millions of independent journalists will be better outcome for humanity.
Ohh. You think being owner of a company whose newspaper is read by hundreds of millions of people every day, doesn't put you in a position of power to control the society?
There is no conspiracy, that's the neat part, it's just how the system itself works.
Media survives through advertising. Those who advertise dictate what gets shown and what doesn't, since if something inconvenient for them gets shown, they might not want to advertise there anymore, which means less money. It's the exact same thing that happens online, it's just more evident online than in traditional media.
How come that even before Oct 7 Europe in general sided more with Palestine than with Israel, whereas it's the opposite for the US? Simple, Israel does a whole lot of lobbying in the US, which skews information in their favor. Calling this "brainwashing" is hyperbolic, but there is some truth to it.
Great. I will start a company to generate training data then. I will hire all those journalists. I won't make the content public. Instead I will charge OpenAI/Tesla/Anthropic millions of dollars to give them access to the content.
> Not paying them their fair share will likely decrease the volume of high quality data available
It won't. Thats not how capitalisam works. If high quality data became unavailable, then companies will be created to fix the problem. Only they look quite different from NYT.
Just like how Torrents didn't kill movie industry. These are lazy arguments made by people who want to make money through lawsuits.
Also I can guarentee you even in worst case, humanity would survive just fine without those high quality content just like it did for the past 50K+ years.
What you should actually be concerned about is stupid law suits like this that can prevent progress.
AI could help humanity solve more pressing problems like cancer.
By getting caught up in silly law suits like this and delaying progress one can make a case that you bring more suffering to the world.
> You do copyright for content that you invented and which didn't exist before.
I dont think that's accurate.
The Copyright Act, § 103, allows copyright protection for "compilations (of facts)", as long as there is some "creative" or "original" act involved in developing the compilation, such as in the selection (deciding which facts to include or exclude) and arrangement (how facts are displayed and in what order).
Not sure where you're coming from in this. A NYT article, once written is copyrighted. Using the content without attribution is at best plagiarism, and spitting it out the way the LLMs do is definitely a violation of if that copyright.
Unless you're telling me ChatGPT has eyes and sources just like the NYT and is worrying events as it sees them too?
I don't understand. So if New York times reported on a new laws of physics and put as an article will became copyrighted? Nobody would be able to talk about it and has to discover it by themselves?
How is reporting on an event different from reporting on discovering a scientific law?
The exact words used to explain the scientific law are copyrighted by the writer (presumably the paper's authors). Rephrasings are not copywrited by the source, but by the rephrasing entity (e.g. the NYT, or a teacher that made a handout for their class).
Copyright on scientific papers is most definitely a thing, by the way.
What makes you think my opinions will change based on how legacy systems work in real world? Just because a stupid system exists doesn't mean it's correct.
It's not. That's why at the end of every article that is not original reporting you will find a little bit saying "As originally reported by (organization)" and there is usually some sort of license associated with that. ChatGPT neither includes sources nor deals with any licensing. That's the issue
Yes. Journalism is a job. People do the work of turning these happenings into words, and are paid for it. That's what's stolen here. The value created through doing that work.
If it didn't have value, Microsoft would lose nothing by no longer ingesting it.
> Does NYT pays money to the people/countries etc it uses to as subject to create content(NEWS)? Isn't that stealing then?
No, that's why in my reply to "facts like happenings in the world are not copyrightable" I emphasised do the work. Journalism is a job. Happenings do not just fall onto the page.
> Also their website TOS didn't prohibit LLMs from using their data.
This is just lazy. We have rule of law. Individuals don't need to write "don't break law X" to be protected by them. And nytimes does in fact have copyright symbols on its pages - not that it needs them.
There is no rule of law saying LLMs cannot be trained on WWW data.
New York times made it ridiculously easy for anyone to access their content by putting it in WWW for making money from page impressions. And they started ingesting links of their content to social media, search engines, etc.
And now they are acting surprised someone used the content to train an LLM.
Should have done their job in the first place to prevent it from training LLMs and make it less.
But they didn't because that affects their page impressions and ad views.
Because the more open the content the more money they make everyone click on a link and see the ad.
You can't have it both ways.
If you do gambling by making content so open so you can get more views from ads, you also get to enjoy the consequences and not cry like a baby asking for billions by making stupid decisions in the first place.
Yeah. I don't think it makes much sense to allow copyright on textual descriptions of events that happened.
Now the question is whether did OpenAI violate the terms of service by using the bits transferred from NYT to train their LLM. I don't think their TOS had LLMs mentioned. So it's on NYT to be negligent and not update their TOS right?
