FTC Gets Fair Use Backwards, Claims It’s Somehow Anti-Competitive?
from the that's-not-how-any-of-this-works dept
This is what I get for praising a move by the FTC earlier this week. It repays me by pushing out a batshit crazy statement regarding AI and copyright. As we’ve been discussing over the past week, the Copyright Office’s request for comments regarding AI and copyright has been leading to some odd comments, including Hollywood saying that it doesn’t want to expand copyright laws (?!?) for AI or the newspaper industry’s trade group highlighting how it wants to demand free cash from AI companies. And, of course, there’s the comment that we filed in the proceedings, highlighting how the right to read should never be blocked by copyright.
There were many, many filings on the docket, and so some interesting ones are still popping up — including a very strange one from the FTC. There are some perfectly fine comments in the filing about protecting consumers (the FTC’s general job), though the FTC has some very weird ideas about how to do that in the AI space, especially as its own filing appears to contradict itself.
Specifically, the FTC claims that it has a significant concern about anti-competitive issues regarding AI, mainly because Big Tech controls so many of the inputs.
The rapid development and deployment of AI also poses potential risks to competition. The rising importance of AI to the economy may further lock in the market dominance of large incumbent technology firms. These powerful, vertically integrated incumbents control many of the inputs necessary for the effective development and deployment of AI tools, including cloud-based or local computing power and access to large stores of training data. These dominant technology companies may have the incentive to use their control over these inputs to unlawfully entrench their market positions in AI and related markets, including digital content markets. In addition, AI tools can be used to facilitate collusive behavior that unfairly inflates prices, precisely target price discrimination, or otherwise manipulate outputs. The FTC is empowered under Section 5 of the FTC Act to protect the public against unfair methods of competition, including when powerful firms unfairly use AI technologies in a manner that tends to harm competitive conditions.
So, almost everything about this paragraph struck me as odd. As we had noted last year, the big surprise in the AI explosion in 2022 was that it wasn’t coming from Big Tech. If you went back a couple years, the narrative presented here by the FTC was considered absolute truth: only a very limited number of companies — Google, Meta, Apple, Amazon — had the data necessary to create powerful AI.
And yet, the generative AI revolution… hasn’t really come from them. Those companies are all playing catch up to smaller companies like OpenAI, Anthropic, and MidJourney. While some of these companies are backed by Big Tech, it’s incredibly notable that the leaders in the field are not Big Tech, and Google, Meta, Amazon, and Apple all look like they’re trying to catch up. If this space was somehow only possible for the Big Tech companies, you wouldn’t be seeing lists like 50 Generative AI startups to watch.
Second, if the FTC is claiming that it has Section 5 authority to deal with this, why is it even commenting here about copyright?
But, the really crazy bit is that if the FTC is correct that the “powerful, vertically integrated incumbents control many of the inputs necessary for the effective development and deployment of AI tools” including “access to large stores of training data,” then you would think, that the FTC’s comment regarding the copyrightability of training data would be to side with the argument that AI systems making use of training data on the internet is fair use.
After all, that would make sure that this data is available for everyone to build AI systems. It would be inherently pro-competitive, making sure that the big incumbents can’t cut off access to training data, or that only the big incumbents could afford the necessary licenses to scan the training data.
You would think.
Instead, the FTC takes the positively ridiculous position that… fair use is anti-competitive. Really.
These liability questions implicate consumer protection and competition policy. For instance, under certain circumstances, the use of pirated or misuse of copyrighted materials could be an unfair practice or unfair method of competition under Section 5 of the FTC Act.
Got that? Apparently making training data widely available to competitors is… anti-competitive. Instead we should require licensing… which only the large incumbents could afford. Because… that’s pro-competitive?
Remember, the FTC has no authority over copyright, and it’s complete and total lack of understanding of the subject of copyright and fair use really shows in this filing.
The use of AI technology raises significant competition and consumer protection issues beyond questions about the scope of rights and the extent of liability under the copyright laws. As the courts apply the doctrine of fair use to the training and use of AI, the evolution of the doctrine could influence the competitive dynamics of the markets for AI tools and for products with which the outputs of those tools may compete. Conduct that may violate the copyright laws––such as training an AI tool on protected expression without the creator’s consent or selling output generated from such an AI tool, including by mimicking the creator’s writing style, vocal or instrumental performance, or likeness—may also constitute an unfair method of competition or an unfair or deceptive practice, especially when the copyright violation deceives consumers, exploits a creator’s reputation or diminishes the value of her existing or future works, reveals private information, or otherwise causes substantial injury to consumers. In addition, conduct that may be consistent with the copyright laws nevertheless may violate Section 5.
This is just a bizarre comment, because what it’s describing in the bolded section is… not a copyright issue? At all? It’s mostly handled by state publicity rights laws. So it’s really strange and revealing that the FTC thinks it’s potentially a copyright issue.
Incredibly, the FTC then admits that the large incumbents have the resources to license data… which might be anti-competitive:
Many large technology firms possess vast financial resources that enable them to indemnify the users of their generative AI tools or obtain exclusive licenses to copyrighted (or otherwise proprietary) training data, potentially further entrenching the market power of these dominant firms. These types of issues not only touch on copyright law and policy but also implicate consumer protection and competition concerns across a wide range of industries
Given that, you would think that the FTC would then be supportive of the idea that training on publicly available data online is not a copyright violation. As that would be inherently pro-competitive, enabling many of those generative AI startups, that the FTC doesn’t seem to realize exist, to compete with large incumbents.
On top of that, if the FTC’s concern (as noted in the quote above) is that the largest companies have the resources to indemnify users… then why isn’t it supporting the claim that the use of generative AI isn’t infringing? Again, that would be a true pro-competitive stance, because it would take away that indemnification power from the large incumbents and would help many of these startups better compete.
The whole filing is… just bizarre. It’s basically saying “the big incumbents have too much power… and we support a confused interpretation of copyright law that guarantees they have more anti-competitive power.”
The FTC has no authority over copyright law, and this filing shows why.
Filed Under: ai, competition, copyright, fair use, ftc, generative ai
Companies: openai
Comments on “FTC Gets Fair Use Backwards, Claims It’s Somehow Anti-Competitive?”
Complaining that “Big Tech controls many of the inputs for AI” is like complaining that “Big Aerospace controls many of the inputs for Air Transportation” because it makes all the big aircraft.
In no small part, folks like Midjourney, OpenAI, etc themselves “became” Big Tech because people like to use comforting categories. They are successful, therefore they are influential, therefore they are Big Tech. Only have a few hundred employees? No matter, we aren’t picky how we apply our labels!
Re:
Big Tech doesn’t own the inputs to AI, they own billions of dollars of hardware, which are needed to learn from the inputs.
Midjourney is not “big tech”, and OpenAI is a shell company for Microsoft at this point, as the recent drama has demonstrated.
I will also highjack this comment to mention that I testified to the FTC about the issue of publicity rights, and mike is correct about publicity rights being completely different than copyright. see e.g. Maloney v T3 Media (9th circuit).
https://www.ftc.gov/system/files/ftc_gov/pdf/transcript-ftc-open-commission-meeting-11.16.23.pdf
Um, boss? The question isn’t “is making the training data available to competitors anti-competitive”. It’s “just who are the competitors you are concerned about?”
I mean, is that mechanized loom anti-competitive to loom makers, or to home weavers?
Re:
They’re saying its anti-competitive, because lots of businesses invested into the idea that “data is the new gold”, and tried to play gatekeeper to all of that data, with the expectation that it would allow them return on that investment.
Then what happened is that Deep Neural Networks trained on the data, and learned the features of the data, so that it can generate lots of new expressive content, and they can no longer act as the gatekeeper to content.
A great example of this, I would argue is Thompson Reuters, and the New York Times, who are mad that people can have access historical facts, without having to pay them for the privilege for doing so.
Dont remember
But,
was it in the USA or the EU that computer code could not be copy written?
When did this change?
Its a good thing to do/say. As the biggest thing of programming is optimization.
You can do the same thing, in many ways with programming. A few changes here and there and its different, and does things differently.
This is like comparing Search engines. they all do the same things, Just differently, and with Different amounts of adverts, and what they find, and WHO paid for what they show you.
The real problem is HOw to make enough money from the internet. you can show only so many adverts. ANd the less restrictions on the internet, may make it free to see but also less PEOPLE needed to monitor it.
To bad, the internet CORPS cant reap money from the scammers. It would pay for lots more, freedom.
Re:
I think this was several decades ago at this point, i want to say in the mid 90s, but I might be hallucinating.
Aaaand just like that...
…the aforementioned pig has crash landed into a prison for the competitively insane.
This comment has been flagged by the community. Click here to show it.
FTC's comments are all good
What the hell are you babbling? All the FTC’s comments were well thought out policy discussion, but the comments you added to it were completely unreasonable extensions of end user powers. It’s like letting end users misuse the products we create and then president’s nuclear arsenal red button would be given to fucking monkeys.
Re:
Did we read the same filing? Not only does the Federal Trade Commission have no reason to venture into copyright (barring some actual anti-competitive/anti-trust issue with how certain big entities exploit their copyrights), and if this is how they’re starting, it’s neither “well thought out” nor “good“.
There’s a HUGE difference between perceived copyright violations and allowing monkeys access to the President’s nuclear codes. One has the potential to cause the end of the world as we know it, while the other just affects the bottom line of a few copyright-loving gatekeepers. I don’t even think the threat of Nuclear War was anywhere near the threat level most sane copyright maximalists say is coming if we don’t stop piracy by (check notes) limiting fair use, which doesn’t even apply to pirates anyway.
Second, it shouldn’t even be a question of if fair use applies. The right to train is the right to read. If you buy a physical book, you are allowed to read it as many times as you like. That in and of itself shouldn’t be a question of possible copyright infringement. If a human is allowed to do it, why is a computer not allowed to? Copyright belongs nowhere near AI, be it training AI or somehow granting copyright protection on AI-generated works.
Nothing about that should be an “unreasonable extension of end user powers”. It’s simply reading the law exactly as it is currently written. And by the way, an end user using your product in a way you do not like is not a misuse or an infringement. No amount of your “babbling” will change that!
This comment has been flagged by the community. Click here to show it.
Re: Re:
Yes. This would be me. I’m a gatekeeper. The world has no access to my technology, until I receive good compensation for the work I do.
The copyright pirate’s operation is like a huge nuclear explosion. It takes away your ability to live in current society, when it explodes on your face.
Its just sad that the good technologies that we spent 10 years developing are stacked to my kitchen cupboard instead of letting end users that actually need the technology to use it.
But we have no choice on the matter. End users can’t find the technology even if we waved the solution on front of their eyes. They just can’t recognize that this is the thing that solves their web publishing 3d adventure problems.
Re: Re: Re:
” this is the thing that solves their web publishing 3d adventure problems.”
Oh, thank god for that. Humanity has been waiting for such a miracle for eons. Now, finally empowered with this new tool, we can push forward into our bright and prosperous future!
Re: Re: Re:2
Yes, that’s the idea.
Re: Re: Re:3
Sarcasm isn’t your strongsuit, is it?
Re: Re: Re:4
Tero Pulkinnen has never had a strong suit. If had a strong suit, he wouldn’t see Meshpage having absolutely no users. The only strategy he’s ever had is praying for the government of Finland to rape people on his behalf.
Re: Re: Re:5
But you have no way of verifying how many users meshpage has. While everyone suspects that there isn’t many, my itch.io user counts paint a very different picture.
Re: Re: Re:6
Except for all the times that you boasted that you had zero users as a means of “proving” how infringement-proof your software is, because not having users apparently means no one can commit copyright infringement with your software.
You yourself claimed that you made only $58 in 10 years, of which $54 were made not by Meshpage users on itch.io but by teaching people you called “idiots” how to code like you.
However many users there are, it’s very clearly not a lot. Otherwise, you wouldn’t be talking about developing “evil mind control strategies” meant to force the government of Finland to use your tech.
Re: Re: Re:7
Sadly this isn’t exactly true. It’s still possible to commit copyright infringement with the software, simply because I use URLs to access content from the internet. Given that internet gods have not deviced an api to check an URL whether license to the content exists that would work with every url on the planet, my software also cannot verify that users are not committing copyright infringement for those base content items. My only option is to put harsh words to the web site’s terms of service and hope the end users actually learn where the critical locations are which need to be verified against infringements.
That said, Terms of Service -system is actually working. People who actually follow those rules, are getting extended verification that the output is not infringing.
Re: Re: Re:8
I mean, that was always obvious, but it took you until now to admit your folly.
Based on the stricter copyright laws that you want everyone else to follow, you must therefore delete all work and files on Meshpage to prevent future infringement.
Re: Re: Re:9
This doesnt mean that the copy protection technologies are useless. Browsers got significant security wins when they implemented this tech. They just failed to apply it to urls, mostly because urls are too visible technology and limiting it would cause riots. But I think they’re doing mistake when they allow illegal urls to exist.
Re: Re: Re:10
Except for all the times that copyright fanatics like you have claimed that they are. If you genuinely believed the copy protection technologies weren’t useless, you wouldn’t be asking for more.
Never mind that browsers aren’t implementing any of this tech. Antipiracy efforts aren’t going after the browser developers, they’re going after the ISPs. Asking browser devs to determine who’s a pirate and who’s not is like asking building managers to design a toilet that automatically prevents criminals from using it by barring people from entering when they place their hand on the door. It’s not practical.
Re: Re: Re:11
This stuff already exists. It’s called:
1) electronic lock to the toilet
2) a camera equipment pointed at the toilet door
3) a security guard with a red button
This technology already exists/is usable/available to anyone willing to use it. Happily we haven’t seen many people take that route, even though the technology exists/is available. It would be sad if criminals needed to urinate to the police station backyard lawn.
But the pirate tracking purposes, your claim is that the technology is impossible to create. And that clearly isn’t true.
Re: Re: Re:12
And nobody uses it, because none of those technologies you’ve proposed actually solves the main issue: it’s not possible for any tech, or human, to identify who’s a criminal on sight and immediately bar them from using the facilities, while simultaneously allowing non-criminals to use it without being inconvenienced.
If it was true, Malibu Media would have used it. They could have done something as simple as watermarking their content, so when their porn got leaked onto the Pirate Bay they’d be able to arrest the incriminating user. But they didn’t even do that, because they thought it would be much more lucrative to fake a website hack and leak and threaten to sue the innocent. Unfortunately for them, the judges put a stop to their little scam.
Re: Re: Re:13
This definitely is not true. When we have information that downloading images from new york times is illegal copyright infringement, we can deploy a check where downloads from new york times web site is forbidden.
Now obviously we cannot forbid all downloads from new york times web site. To get proper copyright protection, we should only allow downloads from user’s own home page domain, but leave it to the user to specify where in the world that home page is at.
Now that we have homepage address, and new york times is offlimits for downloading, next step is to decide when new york times web site is not your homepage. Pirates can give false homepage address (since they don’t follow the law that says that entering false homepage to text element that asks for homepage is illegal), so with false homepage, we need another criteria how to detect if the activity is legal.
But when your software downloads assets from MULTIPLE different url addresses, ALL OF THEM SHOULD COME FROM THE SAME DOMAIN, i.e. if you try to download new york times and boston herald web sites, then something is wrong given that users only have ONE HOMEPAGE ADDRESS DOMAIN. Thus both urls cannot simultaniously come from user’s unique homepage address, since they’re in different domains.
This kind of analysis can detect if arbitrary homepage address specification matches the urls being downloaded. Since homepage address can be anywhere in the world, and user is expected to be able to MODIFY THE HOMEPAGE DOMAIN CONTENT, these limitations are NOT CAUSING PROBLEMS TO ANY LEGAL USERS OF OUR SOFTWARE.
But ILLEGAL USERS requests will be rejected and our software can find large chunk of illegal accesses.
Re: Re: Re:
Ah yes, blaming potential customers for their inability to see just how incredible your shitty product is.
A tried-and-true approach to marketing.
Re: Re: Re:2
Yes. It’s called honesty. Basically you don’t need to keep your happy-happy-face with pepsodent-smile ongoing for long periods of time, you can tell your customers directly that their practices are not up to the standards that we’re used to.
Re: Re: Re:3
You can be honest, but you also don’t have to insult potential customers in the process.
Re: Re: Re:4
Somehow you failed to understand what being honest actually means. If the customer’s process sucks more than what my mom can do with a typewriter, its just honesty that we give information about the failures to the customer. This means insulting them enough that they get the message.
Re: Re: Re:5
I mean, you’ll probably never get the message, but it’s not going to stop anyone here from insulting you.
Re: Re: Re:6
I don’t mind. I already learned how to deal with bullies at age 7 or something. They tried some evil shit afterwards, but that dropped those people from high school and I never needed to see them again. Dumb people get dumber in the system. I never went to the army, so I didn’t get to see how idiotic those people ended up afterwards.
Re: Re: Re:7
So dumb people get dumber in the system, and yet you claim the bullies dropped out from the high school system… so what is it exactly? Do dumb people drop out or get dumber in the system?
Either way, despite you not being in the system, you seem to have succeeded immensely at getting dumber anyway. Only a complete fucking dumbass would think that the government would let you brainwash and murder innocent people in the name of copyright enforcement.
Re: Re: Re:8
Except that brainwashing can happen accidentally without even trying to do the evil stuff that governments can control. You just need to consistently use words until the listener starts to understand the connections between your consistent word usage. Companies that need accurate results have employed these word tricks, but the technology is regularly failing since the activity goes to the brainwashing area instead of the intended efficiency improvements.
Re: Re: Re:5
You’ve never worked in customer support, have you?
Re: Re: Re:6
Tero Pulkinnen couldn’t support anyone out of a wet paper bag.
Re: Re: Re:7
I have explicitly rejected the human aspects of technology development. Gadgets are much more pleasant to work with, since they fail or succeed in consistent manner without taking into account the mood of the person. If your measurement results regularly give random results simply because one of the test subjects have bad day, the obvious solution is to move to area where there is less noise. Thus rejecting humans is warranted, and gadgets are the test subjects of the future.
Re: Re: Re:
And Linus gave his software away, and it proved so useful that companies banded together to form an organization through which to pay him.
Re: Re: Re:2
dogfood has some influence on this phenomenon. When the dogs found out that dogfooding is available, they went and downloaded the kernel. Rest is history.
Our technology is following their footsteps, with our dogfood approach. Now we’re trying slightly larger pile of dogfood and hope that we can every dog from the world to visit our pile and place our 3d engine to all the 1.2 billion web sites on the planet.
While we’re still far for the 1.2 billion target (currently at number 3), our technology has already been hardened against requirements when the whole world starts using it.
Re: Re: Re:3
oh and btw, check https://github.com/terop2/GameApi to see our source repo.
Re: Re: Re:4
Dogfooding would only make sense if you actually used your software in the way a normal user would. Instead what you do is rip off other developers’ 3D models and look for ways that copyright law can be briefly inconvenienced, then cripple your own engine in the vain hope that Chris Dodd might know you exist.
You using github to display your source repo is a surprise, but ultimately a meaningless, insignificant blip. You hate github, and you hate the opinions of others.
Re: Re: Re:5
how is normal user usage patterns different from how I use my software?
This is still within the license that those models give to people like me. My web site displays the original source of the material and the licenses used.
my use of other people’s models hits squarely to the legal usage where license terms are considered and followed to the letter.
Re: Re: Re:6
You use your software purely as a method of displaying existing model files such as mansions and Friday Nights at Freddy’s animatronics. As a software claiming to be on par with Blender or Unreal Engine, it’s absolutely non-comparable.
The stricter copyright laws that you claim to follow do not permit this usage.
According to you, stricter copyright laws do not care. Stricter copyright laws would have required you to hang yourself.
Re: Re: Re:7
It still does not need clone blender’s or unreal’s features. The featureset could be completely orthogonal to whatever primitives blender and unreal has.
This is why you don’t understand what I have built. One of the main features is “a method of displaying existing model files (in your own web page)”. This is kinda poor main feature, given that it has significant copyright problems. When end users are unable to create those models from scratch, their only option is to go to download free models from the web. In our original design, such feature was simply not supported.
But after 10 years of work, even lower priority features are being implemented, and thus we need to take position on features where there exists significant copyright problems. This kind of features need to always be attracting customers to our technology, but our tech can support significantly better features –> we need to move our customers from simply displaying someone elses 3d models towards working 3d model creation processes.
Re: Re: Re:8
Lol, I looked at official blender chatrooms and they’re completely empty of people. Somehow the community has rejected blender.
Check this url: https://wiki.blender.org/wiki/Communication/Contact/Chat
Re: Re: Re:9
You can argue this, but for the average 3D modeler who works in animation to actually use it, some feature overlap does need to exist.
What people don’t understand is why you keep claiming that displaying existing model files in a webpage was the main benefit of using Meshpage. The fact that you call it a poor main feature now is something new, because every time anyone else calls that main feature poor or the like, you keep insisting that the feature isn’t poor, and that “3D websites” will be in demand. Anyone can read up your comment history and realize this. Your entire point is flipflopping whenever you think that it’s convenient for Meshpage. Realistically, it doesn’t make Meshpage a better product. It barely makes it a functioning product. To whit, websites that are “3D” under your definition still aren’t the norm.
You realize that URL you cited literally has links to other places where chat occurs, right? But of course, Tero being Tero, you don’t actually mean to participate, only to mock others for being human.
The truth is that Blender still has magnitudes more users than Meshpage ever will, and you can continue screaming at your government, but it won’t change that fact.
Re: Re: Re:10
Yes, this is because end users have not yet found our technology. It takes a while before new innovations are absorbed by the public. Our parents still wants nothing to do with a smartphone and they still think that sitting on your phone desk waiting for someone to call the landline is the right way to use phones. I don’t expect 3d technology to get any better reception than what smartphones had.
Re: Re: Re:11
Because you chose to advertise on one London bus instead of Finland, where the bulk of your users were supposed to be – but then we know the extent of your attempts to recommend your allegedly child-friendly modeling software to kids. Which is to say, you’ve never done anything of the sort.
If your parents are anything like you, your parents are just as backwater and close-minded as you are and live 15 kilometers away from the nearest human. The fact that they still insist on analog phones instead of smartphones is a pretty big giveaway. People in my parents and grandparents’ generation cannot get enough of smartphones. They’re just as avid users as their descendants.
There’s also the fact that smartphones have plenty of business applications from the mere idea of being able to talk and contact someone on the go. Meshpage… has far fewer relevant use cases.
Re: Re: Re:12
they just had different requirements for their lifes work, when they needed to get electricity working. We just run our computers dependent on stable electricity supply, but never think that there are people who worked hard to get it done.
Re: Re: Re:13
That much is true. You have absolutely no respect for any other human, because you hate having to share a planet with any other living creature. Which is why you can only think of other humans as slaves and wallets to pay you copyright money.
Re: Re: Re:14
They are more like empty shells of gray brain matter. Once you’ve learned category theory, the ivory tower gets so bad that you start to think other people as if they’re idiots who cannot do their jobs properly, simply because they don’t have the theorethical groundwork necessary to understand the fine aspects of logic and functions.
The copyright aspects are clear consiquences of the above. When the humans cannot recognize how important copyright’s finer details are, they tend to overestimate the allowed area and go to the illegal grey areas that are just turning red.
Courts are laughing like hell when they hear the arguments forwarded by the copyright criminals. Fair use has explicitly been put to illegal area, simply to avoid the situation that the people didn’t know their activity might be illegal when they try to get fair use accepted by court system. These criminals are stretching the allowed area, trying to get some advantage over other vendors in the market by working in the gray market. It’s plain old corruption what is happening in fair use area.
Re: Re: Re:10
Yes. When this 3d engine in meshpage was implemented, I was very much against introducing support for .obj, .stl, .glb, .gltf files, simply because creation of those files isn’t possible with the tools available to us. It’s just if internet happens to have those files, are we able to open them. The sad thing about .glb/.gltf is that khronos had designed it properly, and as a end result is that its actually good feature.
It’s poor feature since copyright story isn’t correct in it:
1) you can’t claim ownership of the files if you didn’t create the file yourself
2) tools do not support creating the files from scratch.
3) it’s unknown how the existing files from the internet are being created
The problem is that there is significant demand for the support of these file formats, given that internet is full of .obj and .stl files. Demand for one file format doesn’t need to translate easily to demand to our 3d web plans.
Re: Re: Re:11
That’s a whole lot of text just to say “I hate that I have to work with other file formats, solely on the basis that I want people to use MY exclusive file formats instead, and I will assume that anyone who disagrees with me is a pirate and a criminal.”
Your loathing of other humans knows no bounds. And you wonder why the government of your country hasn’t made you their state religion yet, when you continue to be the most antisocial boomer I’ve ever seen.
Re: Re: Re:12
Lets analyze your claims once more;
1) These other file formats give access to internet’s content collections. I.e. the pirate sites.
There is better solution than this. Instead of supporting common file formats, you could create something original, i.e. make the software implement features that are not supported by the existing file formats.
2) It’s this “originality” requirement from copyright laws that you definitely do not understand.
Yes, these people in (1) have access to pirate sites, and there’s only small step from that to actually committing copyright infringement. The people who don’t understand (2) have trouble also understanding how “Promote the progress and useful arts” actually works. Replicating work that someone else had already done, or copying their work is not as valuable activity than creating original copyrighted work from scratch. This is because the writings shouldn’t be the same message duplicated millions of times and letting the same message get stronger and stronger until noone can withstand the brainwashing. When different message is conveyed, it doesn’t have the same brainwashing problem than what happens by copying the same stuff over and over again.
This comment has been flagged by the community. Click here to show it.
Re: Re:
Well humans are not allowed to do it either. The cases we’re dealing with is things like passing pirated material from pirate sites to the AI to monetize the illegal activity. Same material given to humans would require the human to reject the material as copyright infringement before learning it. Failing to do the reject, and internet community will bash the person to death by wooden sticks simply because they recognize the original source material to come from pirate sites.
Re: Re: Re:
Massive citation needed there. Is there evidence that they’re sourcing stuff all from pirated material and not, oh, I don’t know, content that is first posted online that can be accessed by anyone for free anyway? Unless you’re prepared to say that Google doing the same thing to build a search engine is infringement, existing case law notwithstanding, then there’s no law broken here.
This comment has been flagged by the community. Click here to show it.
Re: Re: Re:2
Citation below:
https://www.techdirt.com/2023/09/26/authors-guild-jealous-of-other-terrible-author-lawsuits-against-ai-decides-to-join-in-the-party/
says explicitly:
“97. Some independent AI researchers suspect that Books2 contains or consists of ebook files downloaded from large pirate book repositories such as Library Genesis or LibGen,” “which offers a vast repository of pirated text.”
“Other possible candidates for books2’s sources include Z-library, another large pirate book repository that hosts more than 11 million books, and pirate torrent trackers like Bibliotik, which allow users to download ebooks in bulk”
Re: Re: Re:3
A suspicion is not evidence of it happening.
Re: Re: Re:4
evidence was not required. Its enough that “plaintiff’s claims are accepted as true”, so even if no real evidence is forthcoming, this suspicion can be used to financially ruin the defendent.
Re: Re: Re:5
“Plaintiff’s claims are accepted as true” simply refers to the standard for a Motion to Dismiss. The idea being, that even if the court was to assume that the Plaintiff’s accusations were true and bared out in discovery, if there still is no cognizable legal claim that a court can grant relief for, then the case can be dismissed. If the case is not dismissed, then the case proceeds to an answer and discovery, where evidence must be gathered to support or refute the claims made in the lawsuit. That presumption no longer has any legal weight. So, no, simply saying you “suspect” something in a lawsuit, and it’s accepted as true for the purposes of evaluating a motion to dismiss, is not in and of itself evidence that said suspicion is true.
Re: Re: Re:6
Yes, but when this happens, there’s 2 things that are better for plaintiff (not just one):
1) plaintiff’s research on defendants actions is finished.
2) defendant can no longer afford paying for his attorney.
So this means that plaintiff no longer needs to use money for the case, and also the defendant is losing the only thing that keeps him out of the jail cell.
Re: Re: Re:7
Neither of which suggest that there was any good reason to believe that any laws were violated or that any laws were actually violated, so I fail to see how that helps your case.
Re: Re: Re:8
This is why you should have settled the case when you had chance and the lawyer’s bill was yet not that large. Settling would mean you need to admit being a filthy criminal like you are. Content industry rather get quick wins than litigate the issues to its end.
Now that lawyers have burned their allowance, it’s very difficult to get rid of the lawsuit, given that content industry will have to argue endlessly that you’re the filthy criminal, even if that wouldn’t be the case in your la la la land inside your brain.
But if you make single mistake in your paperwork, your case will be decided to your failure and plaintiff is the prevailing party in the case.
Re: Re: Re:9
Well, that much was already obvious. Unfortunately for you, after Prenda Law and Malibu Media tried that strategy one too many times, the judges have finally decided that they’ve had enough of acting as your money mules and now they demand actual evidence.
That era no longer exists. Even the RIAA has decided that they’re not in the business of suing end users anymore, certainly not in the magnitude of their heyday.
Fucking up the paperwork is precisely why so many copyright plaintiffs can’t get their cases off the ground, just like the plaintiffs in the AI-generated art lawsuit right now. Their lack of copyright registration really came to bite them back in the ass and the judge wouldn’t even let the case move forward.
Re: Re: Re:10
Legal area is special in the fact that the paperwork needs to be top notch or people’s lives will be ruined by the sloppy practices. Titanic didn’t sink because their paperwork wasn’t right, but instead they designed it correctly, and was well aware what conditions need to be fulfilled before it sinks to the bottom of the sea, but Titanic sink because they simply couldn’t avoid those situations where the boat would eventually end up eaten by fish.
Whether you have evidence or not simply doesn’t matter, if avoiding the illegal practices are not possible or available to the people involved. Their titanic will sink even if they quite well know under which conditions sinking is plausible outcome. Having evidence of the fact does not change the outcome.
If you want real change, you should do research on alternative approaches. Instead of going to whatever locations markets are leading you to, you should go against the flow and find areas of interest which are getting less support than required, and then give those areas the support they need.
This is what my technology is doing. Instead of cloning blender and unreal, we went against the flow and found out that web publishing area is missing 3d models. Our 100 million phones without opengl probably killed the technology, given that at the time, it was more important to get your web sites to work in mobile environments. To fix the blunder, our technology is now ready and after 10 years of development, we can now say that we have not broken anything in the world while the world was our responsibility. Now its end user’s task to embrace the tech and not leave it unused.
Re: Re: Re:11
You spent an entire tortured analogy to try and argue that paperwork isn’t important, but I should have expected no less from someone who claimed that the Russo-Ukrainian war started because Putin was fighting in the name of copyright law.
lol, and how did that turn out for you? Meshpage supposedly started out as a 3D modeling software for kids, because you claimed that adults were too stupid to understand how simple your user interface was. When you were asked if you were actually teaching kids to use your software, you backpedaled and argued that Meshpage was for college-level students.
Of course the truth is that you’re not serving any market, especially considering that you think all other humans are idiots, and keep trying to leverage a claim of “$54 made in 10 years” to justify raping grandmothers in the name of copyright law.
Nah, it really isn’t. No country will ever make it illegal to not use Meshpage.
Re: Re: Re:12
It doesn’t need to be illegal, just inconvinient enough that they will rather use meshpage than skip the next evolution step we’ve planned for computers.
Re: Re: Re:13
Youtube has just showed how it can be done. Google’s engineers are adding 5 second delays to all videos if you use adblock, this means that the old video-based technologies are very cumbersome to use. Our 3d graphics can skip the artificial 5 second delay and thus result in smoother user experience.
Re: Re: Re:14
It’s been, what, ten? Fifteen years since Meshpage was worked on? In that amount of time you’ve done nothing to spread the name of Meshpage besides whining on Techdirt, maybe Github. Even then the bulk of your contributions has been screaming that we’re all going to be arrested by the RIAA for not using Meshpage, which is something that is realistically never happening because the RIAA doesn’t care about a random asshole in Finland.
And Google is losing that fight. Not just because adblocker developers are constantly updating their software, but what Google is doing is very likely a violation of GDPR and other EU privacy laws.
Just like every other website that doesn’t rely on your supposed technology.
Re: Re: Re:15
Not all of these have even chance of replacing youtube. My technology has all the pieces available, just the popularity is missing.
Re: Re: Re:15
Where you get your numbers from? My github repo says the last change was “last week”. (And I was sick the whole week, so no time for development). How you turn “last week” changes to the “fifteen years” is kinda anyone’s guess?
Re: Re: Re:16
Not every website exists for the sake of replacing YouTube. Not every website’s purpose is to host videos. The fact that a website doesn’t replace YouTube is a moot point that doesn’t justify revamping the website with Meshpage tech.
You started Meshpage anywhere from ten to fifteen years ago. I don’t mean to claim that Meshpage wasn’t updated since ten to fifteen years ago. It’s not my responsibility to make up for your shitty understanding of the English language.
Re: Re: Re:17
Google has built technological barriers to protect against this problem.
Replacing youtube does not (and should not) host videos. It¨s yesturday¨s technology and google had a good run with it.
There is technological barriers built by large companies to prevent this:
1) twitter doesn’t allow embedding 3d models, but embedding videos is supported, and thus 3d models need to be behind url link, and users need to show their interest in the material before seeing the actual content.
2) facebook has similar problems, but the tech works slightly differently
3) linkedin has the same url link problem.
Re: Re: Re:
” passing pirated material from pirate sites to the AI to monetize the illegal activity.”
I must be in need of more coffee.
Why would AI need to go to pirate sites in order to train upon said material when said material is readily available everywhere.
Also, does AI train on music?
Re: Re: Re:2
Everyone else is…. tadaa…. following the law…
and thus they only have small amount of content available. Its legal requirement that platforms that offer content to the public are….tadaa…. licensing the material from content owners…
Re: Re: Re:3
And just how many ebooks can the likes of Google buy using a petty cash account?
Re: Re: Re:4
if google has 2 billion cash available for licensing, they can support 2000 artists with 1 million bucks each. RIAA and MPAA will give google an award of best supporting actor when that happens.
Re: Re: Re:5
Just because you attempt to slap someone with a fine, it doesn’t mean that person has the money you claim they have.
Which is why the RIAA’s campaigns failed so consistently; they kept demanding money from innocent people that those people never had.
The RIAA and MPAA would sooner throw themselves off a cliff than give Google anything.
Re: Re: Re:6
You have to understand RIAA’s playbook. It doesn’t matter how much BS the member companies are receiving from the public at large, when the other side of the scale is some pirate operation which is stealing from the actors, filmmakers, musicians and directors. Those people who anyway didn’t have more than 2 billion bucks for signing up to the serial adverticements every half year and getting money flowing simply by the actor’s face visible next to some god awful product that sells badly without lightsabers.
Re: Re: Re:7
You would think so, and the RIAA thought so at one point. They genuinely believed that the law was on their side and they could sue children, grandparents and corpses with little to no evidence. Eventually public opinion and backlash on them was so bad though that they never went any further despite winning the Jammie Thomas-Rasset and Joel Tenenbaum cases. Unfortunately for the RIAA, they won two lawsuits but absolutely lost the reputation and goodwill they had among the public. Nobody other than the most insane of copyright fanatics believes that downloading a CD should carry a fine that’s much, much larger than what would have been charged if the CD had been shoplifted instead.
Those people have already been paid for work up front. You realize that, right?
Guy, have you even watched a movie that isn’t Star Wars? I’ve paid the MPAA several times last year to watch other films. None of them featured lightsabers. Maybe you should get out of your parent’s house for once and go touch some grass. Or touch some snow, or whatever the fuck is out there in that Finnish wasteland you’ve camped in.
Re: Re: Re:8
Your idea of how the money is calculated in copyright infringement cases is completely broken. Correct version goes something like this: It takes $300k euros to create the original work. When that is split between 100 users, each user need to pay $3k euros to license the material. The copyright infringer uses the material without the “discount” step, so each infringer should pay $300k for the “permission” to infringe. This is about what damage awards courts have declared against the copyright infringers.
Re: Re: Re:8
Well, the money that RIAA/MPAA gets from copyrighted work gets invested to creating new copyrighted works. You realize this? If the money is coming from other sources than copyrighted works, then authors are not receiving that part of the money, since its needed in whatever area generated the money.
Software area especially is dangerous, since on internet no one wants to pay for software- The money goes to purchasing hardware and necessary items for system to work, but software area is expected to be delivered for free.
Re: Re: Re:8
Joel’s argument was that piracy is like speeding in the information superhighway. So he clearly admit illegal stuff, given that speeding is also illegal. Of course RIAA is going to win the case in such circumstances.
Thomas-Rasset was some single mom who didn’t know who is using her internet connection?
Re: Re: Re:3
“Its legal requirement that platforms that offer content to the public are….tadaa…. licensing the material from content owners…”
Are you an attorney providing legal advice on a tech blog?
Are you entirely sure about what you stated?
Re: Re: Re:4
It’s terop, you can safely assume anything he says has little to no connection with factual reality.
Re: Re: Re:4
No, I don’t need to be a legal professional to know that there’s something fishy about how copyrighted works are handled in the marketplace.
Yes, the licensing operation has falled out of favor from the market participants, since you actually need to move money around to do it. Everyone looking at their own bottom line, there simply isn’t any money reserved for mundane operations like licensing the content that your product uses.
In fair world, those products which do not pay content owners the money the authors deserve should release their products without content to the marketplace. But do you see lots of popular products in the marketplace without any content items available in the product? No you dont, and the reason is that users are screaming to get their hands to good looking content. But the middleman just don’t want to pay to produce it.
Re: Re: Re:5
Really? Because it sure doesn’t seem like it from my perspective. If anything, licensing seems to be trying to replace actually selling the product.
Re: Re: Re:6
that was happening in 1980-1990 timeframe, but the time for that is already gone.
Re: Re: Re:7
Aw, too bad, so sad for you.
We’re not going back to the supposed golden age of the 1980s to 1990s.
You’ll just have to deal with it by being not successful.
Re: Re: Re:8
You get this kind of problems when you fail to embrace our technology. Reviewers have praised it for 10 years of being so 1990s, and we’re sure we’ll reach full 1980s experience any day soon.
There are many different levels of success. And to get our software ready, we’re willing to suffer various levels of success.
Re: Re: Re:9
You don’t have reviewers, Tero, because you don’t have users.
And even if you did, none of them is praising it on the basis of reminding them of the 1980s. Retro approaches only work so far as aesthetics, not functionality.
The only “success” you’ve managed to suffer is making only $58 over 10 years. I know that is the only success you’ve had because you won’t shut the fuck up about it.
Re: Re: Re:10
Money isn’t the only deciding factor, we have
1) 600 features the in engine and builder tool
2) 2^{600} feature combinations in feature list
3) k*2^{600} feature combinations after handling type checking limitations
These are all very good successes.
Re: Re: Re:11
Which you’ve also boasted about hiding behind a paywall and/or removing in case they might be used for imaginary copyright infringement. You could boast about having infinite features and it wouldn’t matter a damn thing, because anyone can boast about things that could happen, but don’t.
Also “k*2^600” doesn’t mean a thing if k=0.
Re: Re: Re:12
There’s no paywall. While I ask for money, the download is free and it’s up to the honesty of my customers whether I get any money for the work I’ve made. The software works perfectly well even without involving payment system to it and the payment is only required if you actually want/need a license to the material.
Re: Re: Re:7
I’m looking at the 2000-2020s. People have to license their phones and tractors nowadays.
Re:
So, those were some words.
The real cost of the generative AI is not to find the data (even filtering to keep only the relevant data) but the computing power needed to train theses models.
It take around 2 weeks to train a chatbot of OpenAI, and every request a dozen of (computing) seconds to proceed, which translate to hundreds millions dollars a year for a moderate successful service.
Microsoft have a great amount of computing power available with Azure (because not enough clients). Same goes for Amazon with AWS. And Google and Facebook have some large datacenters to play with.
A startup wouldn’t have more than few millions bucks to start with theses hungry beasts, and so, the results would be a irrelevant as it was few years back. OpenAI is bleeding money badly, and without the (financial) help of Microsoft, wouldn’t be there no more.
Since it seems that the return of investment is close to zero (who gonna pay for “Copilot” AI in Windows?), only big players can pay infinite bill until they realize, just like for this “metaverse”, that this “revolution” is mainly based on hype. They’ve go the marketing to impose such a technology to people, but may not enough to impose to shareholders.
Re:
My technology at meshpage.org has exactly this problem. The end users are not ready to embrace the newest technologies, and there’s danger that our software platforms are getting old/rotten before end users are ready to explore the new technology. Our tech lets users play with 3d creation toolchains, and post the end result to a web site of their choice. This kind of features are only currently available as short text snippets or jpg/png images and limited amount video files. But 3d models and more advanced graphics primitives are missing from the available tools. This gap in the technology landscape must be corrected before end users start creating their platforms and services. They need to have experience of what 3d modelling is like, so that they can embrace the technologies in the next chapter of our technological marvel.
Re: Re:
It’s not the responsibility of others to give you attention and engagement before they start designing their own solutions.
Re: Re: Re:
If they fail with the attention and engagement department, then they will just lose existing working solutions and be endlessly recreating the same stuff, simply less stable software and less accurate timing. It’s going to end in tears. Now the technologies are available and working, and we’re simply improving the documentation aspects. For example, I created this API reference yesturday:
https://ssh.meshpage.org/mesh_doc.php?menu=0&submenu=0&select=select&1831127721
Re: Re: Re:2
The failure with the attention and engagement is on you, because your only avenue for attention and engagement is a bus in London who ran one ad from you in one week. You’re not getting attention because you’re not talking about your engine at all besides trolling on Techdirt to a disagreeable audience.
And the target audience you actually want, who you’re hoping will pay you money for Meshpage, doesn’t even know you exist. They’re not losing “existing working solutions” because they’re not giving you attention. They’re still going to be using 3dsmax, Maya, Blender, Pixar, etc for their purposes.
The market demands for more of the same things, so no – I doubt it’s going to end in the tragedy you desperately imagine just to vindicate your crappy bloatware.
And how much did you pay Oracle to use the term API?
Re: Re: Re:3
I have pretty much tried all these (except Pixar) before deciding to create a tool of my own. 3dsmax and Maya failed for onerous requirements for licensing in the installation app. Blender was just too difficult user interface for me. Pixar I haven’t seen/heard about.
Re: Re: Re:4
Sounds like a skill issue.
Re: Re: Re:4
I visited your webpage, and I want you to know that its absolute garbage, and that you would have been better off just using the quake engine and compiling it with enscripten into webassembly, because at least it would have actually ran performant (while still looking as dogshit).
This is probably one of those situations like Terry A Davis, where you decided to roll your own 3d rendering pipeline, despite the fact that it takes several hundred man years to make a GOOD rendering engine, including all the work of computer graphics PHD researchers.
Even the best 10X engineer should not endeavor to roll his own.
Re: Re: Re:5
Yes, but this choice would limit us to running quake maps, which have been declared outdated and not original already in 1995 or something.
Our current solution does still result in outdated stuff (because we create it ourselves), but at least the originality aspect works better when you pull a rabbit from your hat instead of borrowing other people’s work.
Re: Re: Re:5
This is similar situation than if local food supply has a shopping mall available, and then new corn producer introduces a machine that can create corn flakes to the market. The shopping mall owners will dismiss the work simply because it cannot do 1200 different products that their shopping mall is able to sell to the customers. But the people who introduced the corn flakes machine sees it differently. Instead of trying to fill needs of one person who visits the mall, the corn flakes machine tries to fullfil the needs of multiple customers who all want good breakfast.
We don’t need a shopping mall, when our engineers can provide good breakfast for the whole country.
Re: Re: Re:6
And thank goodness engineers aren’t in charge of agriculture or food production or marketing.
At least, not engineers like you, given that you’ve proven time and time again that you don’t give a shit about what “good breakfasts” the consumers actually want.
Re: Re: Re:7
consumers themselves do not seem to know what they want, and that’s why engineers need to explore different areas to find best feature combinations for end users. After such exploration, my web site can be offered to customers as examples of what is possible with the technology.