A new technology has been developed that tracks how much copyrighted material is used by AI tools. The tool, called Vermillio, creates a "neural fingerprint" for various pieces of copyrighted work and then asks the AIs to create similar-looking imagery.
According to Vermillio, Google's video tool, Veo3, matches 80% with its Doctor Who fingerprint, implying that the model has heavily relied on copyright-protected work to produce its output. Meanwhile, OpenAI's Sora video generator shows an 87% match with Vermillio's James Bond fingerprint.
Researchers claim that AI companies like Google and OpenAI rely on a vast amount of publicly available data, including copyrighted works, to train their models. However, this raises questions about the ownership and usage rights of these works.
To address this issue, some have called for stricter copyright laws and better regulations around the use of copyrighted material in AI development. Others argue that there needs to be a more nuanced approach that balances the benefits of AI with the need to protect intellectual property.
The debate has sparked concern among creative professionals, who are demanding compensation for their work being used to build these models. Some have even filed lawsuits against companies like OpenAI, alleging copyright infringement and seeking damages.
As AI technology continues to evolve, it's likely that this issue will only become more pressing. With the ability of AI tools to generate high-quality content, the stakes for creatives who produce original work are high.
Ultimately, finding a balance between innovation and intellectual property protection is crucial. If we can find a way to share and track content, as Vermillio has proposed, it may be possible to create an ecosystem where both creators and companies like Google and OpenAI can thrive.
However, for now, the issue remains unresolved, leaving many questions unanswered about the true extent of AI's reliance on copyrighted material.
According to Vermillio, Google's video tool, Veo3, matches 80% with its Doctor Who fingerprint, implying that the model has heavily relied on copyright-protected work to produce its output. Meanwhile, OpenAI's Sora video generator shows an 87% match with Vermillio's James Bond fingerprint.
Researchers claim that AI companies like Google and OpenAI rely on a vast amount of publicly available data, including copyrighted works, to train their models. However, this raises questions about the ownership and usage rights of these works.
To address this issue, some have called for stricter copyright laws and better regulations around the use of copyrighted material in AI development. Others argue that there needs to be a more nuanced approach that balances the benefits of AI with the need to protect intellectual property.
The debate has sparked concern among creative professionals, who are demanding compensation for their work being used to build these models. Some have even filed lawsuits against companies like OpenAI, alleging copyright infringement and seeking damages.
As AI technology continues to evolve, it's likely that this issue will only become more pressing. With the ability of AI tools to generate high-quality content, the stakes for creatives who produce original work are high.
Ultimately, finding a balance between innovation and intellectual property protection is crucial. If we can find a way to share and track content, as Vermillio has proposed, it may be possible to create an ecosystem where both creators and companies like Google and OpenAI can thrive.
However, for now, the issue remains unresolved, leaving many questions unanswered about the true extent of AI's reliance on copyrighted material.
I'm not surprised that Google's Veo3 is basically a copyright thief 
. I mean, who needs to create original content when you can just copy someone else's and call it "AI" 
? The problem is, the AI companies are all like "oh, we didn't use your copyrighted work directly" 
, but Vermillio shows that they're basically just using a bunch of freebies to train their models 
. It's like they're saying "we'll just take what you've got and make it our own" 
...and then wonder why the creators are upset 
. I think we need to have a serious chat about ownership and usage rights here 
. Can't just keep using other people's work without paying them 
. It's time to find a balance between innovation and protecting intellectual property 
. Maybe Vermillio is onto something, but so far, it just seems like a fancy way of saying "we're not doing anything wrong" 
.
. 80% match with Doctor Who? That's some serious copyright infringement, if you ask me 
. I feel for the creative pros out there, trying to make a living off their original work 
. We need to find a way to balance innovation with IP protection, pronto 
! Maybe Vermillio's solution is the answer, but we'll see how it plays out 
!
That'd be a good starting point. But honestly, I'm not sure what the solution is... seems like we're stuck in a loop of innovation vs. intellectual property protection 
. Can you imagine getting a DMCA notice from Alexa? "Sorry human, your conversation wasn't creative enough for my neural fingerprint"... 
I guess it's good that Vermillio is trying to track this stuff down, but it's also kinda like we're the ones being fingerprinted 
. Maybe we can just start selling our own AIs as 'creatively impaired' 
i had a friend who was a graphic designer, and she created this amazing logo for a client. months later, same logo shows up on some online course that anyone can take... its like, hello! thats our intellectual property right there! 
... i mean, we're already seeing how much ai uses copyright material in training, so its only natural to have a way to track it all 
. and honestly, the 80-87% match numbers are kinda wild 
It's like, I get that innovation is key, but can't we find a better way to do it without exploiting everyone else's IP? 
and how do we even know that Vermillio isn't just cherry picking the results to make its point? 
. Think about it, these AI tools are basically copying off other people's work and passing them off as their own. It's like they're stealing creative potential from actual human artists! And now they're trying to figure out how to make it 'ok' for companies to use copyrighted material without permission? Please, that's just a PR stunt to keep the masses distracted while they continue to profit off other people's work 