Copyright Apocalypse: Trolls Attack the Net, From the Future

HardOCP News

[H] News
Joined
Dec 31, 1969
Messages
0
I've read this article twice now and I still have no idea how this is even possible.

Imagine an Internet in which every possible creative work uploaded results in a copyright claim - because it's already been created. That's the nightmare scenario being painted by a Russian company which says it has a plan to use copyright and trolling to free humans from ever having to create digital content again.
 
What a fantastic idea to prove how idiotic and worthless our copyright law currently is.
 
There's two fundamental flaws in this scheme, at least under US law:

1. To obtain a copyright, your work has to be original and fixed in a medium. It does not have to be novel. If they produce an 1000 x 800 image and I produce the very same 1000 x 800 image without copying their image, both images can be copyrighted. There's no first to copyright requirement -- simply a requirement of originality to the author.

2. To obtain damages for copyright infringement, you have to show that someone had access to your work and more likely than not copied it. Similar to the originality requirement, if you mathematically generate every 400 word paragraph possible and I produce the very same 400 word paragraph, I do not infringe the copyright unless I've copied your paragraph (from you or from another source). There's no there-can-be-only-one ownership right -- rather I must have actually copied (more likely than not) your accessible copy of the work.

Generating a massive and closed database of images and words gets you nowhere. Generating a massive and open database of images and words only gives you the thinnest basis to allege that someone copied their image or their words from yours. As you mathematically exhaust all possible combinations, the naive likelihood of someone having copied from that database paradoxically goes down. The odds of you generating that paragraph and me having generated that very same paragraph -- multiplied together -- mean that the odds of the two events having happened independently are very rare. The odds of you generating that paragraph upon having generated EVERY possible paragraph times the odds of me having generated any particular paragraph -- multiplied together -- mean the odds are substantially higher. The same as the odds of having generated the paragraph independently in the first place.

As a result, the troll will have to prove forensically that the paragraph was copied from them, not merely that they have the same paragraph in their database.

They'll troll threatening the expense of a suit. Experts will develop a fill-in-the-blanks kit for a motion for summary judgment based on the lack of forensic evidence of copying. Judges will rapidly become annoyed. It will not happen quickly, but it will happen if they have the gall to try the scheme. And following in the steps of Rightshaven and Prenda Law, it will end badly.
 
There's no there-can-be-only-one ownership right -- rather I must have actually copied (more likely than not) your accessible copy of the work.
So how does that work so far with companies suing others ones because they claim to have the rights to it? It's not a matter of actually copying, it's a matter of you being the first therefore no one else can profit from that. And yes you can sue, I seem to recall Harley Davidson got a copyright or trademark for the sound their motorcycles make, and yeah they did sue others who used pipes of a certain diameter that just happened to be a standard size, that would make that same sound.
 
http://ktetch.co.uk/2014/09/qentis-business-plan-doesnt-add-up/

Quentis claims to be generating “all possible images in the 1000×800 pixel format”. The article above does the math, and shows that even generating all possible 1000x800 black and white pictures would mean 9.920 x 10 to the 240823th power pictures.

It would take them billions of years to do so, and our universe doesn't have enough storage space to hold all those pictures.
 
hmmm, in that case they might be better off turning their resources to cracking bitcoin addresses
 
Its a good intellectual exercise but little else. In a way they actually invalidate all their own claims to copy right because they admit to doing nothing other than randomly generating all combinations of words and that actually suggests that no one at all would have actually bothered to copy their works since it would be impossible for anyone to read even a small fraction of them to begin copying. Second they have not freed any corporation from having to pay anyone anything because they have not proposed any possible way that a corporation can possibly go through their entire database and effectively pick out the work of art that best suits their needs without hiring an artist to select the work but of course that artist has no capability to read even a fraction of that library so in the end it ends up being cheaper to just recreate a new work from the artists mind independent of their stupid library. In addition making a copy right claim to those works will surely land generate existing works which they might end up liable for infringing on.

This is kind of funny because these guys are actually arguing the exact opposite of what the prevailing argument is becoming. In many fields data generation is becoming so verbose and so deep that the problem isn't generating data its actually analyzing the data and storing it or doing anything useful with it. Its becoming so expensive to do both of those things that people are starting to consider throwing out perfectly good information and data because its actually cheaper to regenerate if they need it again in the future than it is to store it and analyze it all. Consider having a record of information on every human on earth, then realizing you don't know how you could possibly make any use of that. So you spend 15 years developing an ability to screen that data and figure out where to put coke machines. By the time you finish and analyze the data its all old, 10% of the population in the data is dead, 20% have moved to different locations and all the money you spent storing the data has eaten up the efficiency you gained putting coke machines in better places.

As another poster said, do the math you cant even afford to store all these images on drives with the money you would generate. This is all about efficiency and until you algorithm can learn to throw out all the garbage and way to close duplicates the data overload will drown you.
 
Back
Top