🚩 Report: Copyright infringement
Most peoples of Cara do not appreciate the scraping of their site. This repositories existence is actively causing well-being (and physical as a result of that) issues for many members who want to own their drawings as a creation of theirs protected as a work of theirs, and strongly oppose AI and do not want to be involved in it. It is true that some artists do not care, but the opposite is also true, and people should be asked at the very least, as some data is legitimately collectable from Cara from people who submit Public Domain licensed work or works with other licenses that permit AI use.
https://smallbusiness.chron.com/urls-copyrighted-62319.html
Copyright and URLs
URLs are not protected by copyright. This allows others to reference or link to a website address without needing to get permission from the site's creator. Legal disputes over domain names and URLs are not resolved under copyright law but instead by authorities such as the Internet Corporation for Assigned Names and Numbers (ICANN).
Good point! It is true that indeed the dataset holds URLs that link to images from Cara's CDN.
The URL itself existing and being shared is not a problem, it is what let's us access the image that was uploaded (and compressed) by Cara to the platform publicly for viewing.
However this is a dataset, a collection of data taken from a single source or intended for a single project (per the definition from Merriam Webster).
With this we believe that hosting a discovered amount of URLs from Cara, a site for Artists that are known to have anti-AI members, and posting them onto an AI hub website ("The AI community building the future." HuggingFace's about page), has been done in bad action or faith to encourage real copyright infringement if not just general online harassment of an entire community.
We also believe there is no other reason for this datasets existance other than for inevitably scraping Cara's images from a now easily accessible list of URL's.
No one on huggingface is legitimately going to explore Cara images this way ("explore" as in view and exit with no further action other than inspecting details). Any person who would do so are fine to view these images, but this is not the datasets intent, especially after seeing comments from specific pro-AI persons who appear to be wanting to scrape Cara images by listing their URLs to then be scraped.
https://cara.app/post/d60e947c-fd54-45ba-a44b-101fbf8069e3
If cannot be true that this evasion of hosting URLs to content on Huggingface magically voids the copyright of the content it links to, this is the problem.
From the same page you linked to me:
Copyright
Creative works such as books, paintings and photographs are protected by copyright law. Once an artist or author creates a work, copyright gives the creator the sole right to its use, with some exceptions. For example, students can quote a small phrase from a written work without violating copyright.
I am not against discussion, if you would like further clarification on a certain point, feel free to let me know!
Good point! It is true that indeed the dataset holds URLs that link to images from Cara's CDN.
The URL itself existing and being shared is not a problem, it is what let's us access the image that was uploaded (and compressed) by Cara to the platform publicly for viewing.
However this is a dataset, a collection of data taken from a single source or intended for a single project (per the definition from Merriam Webster).
With this we believe that hosting a discovered amount of URLs from Cara, a site for Artists that are known to have anti-AI members, and posting them onto an AI hub website ("The AI community building the future." HuggingFace's about page), has been done in bad action or faith to encourage real copyright infringement if not just general online harassment of an entire community.
We also believe there is no other reason for this datasets existance other than for inevitably scraping Cara's images from a now easily accessible list of URL's.
No one on huggingface is legitimately going to explore Cara images this way ("explore" as in view and exit with no further action other than inspecting details). Any person who would do so are fine to view these images, but this is not the datasets intent, especially after seeing comments from specific pro-AI persons who appear to be wanting to scrape Cara images by listing their URLs to then be scraped.
https://cara.app/post/d60e947c-fd54-45ba-a44b-101fbf8069e3
If cannot be true that this evasion of hosting URLs to content on Huggingface magically voids the copyright of the content it links to, this is the problem.
From the same page you linked to me:
Copyright
Creative works such as books, paintings and photographs are protected by copyright law. Once an artist or author creates a work, copyright gives the creator the sole right to its use, with some exceptions. For example, students can quote a small phrase from a written work without violating copyright.I am not against discussion, if you would like further clarification on a certain point, feel free to let me know!
The URL itself existing and being shared is not a problem, it is what let's us access the image that was uploaded (and compressed) by Cara to the platform publicly for viewing.
Stopped reading there. The images are being accessed by the browser by sending a request to the Cara URL. None of it is being cached by HF's CDN or the likes.
The URL itself existing and being shared is not a problem, it is what let's us access the image that was uploaded (and compressed) by Cara to the platform publicly for viewing.
Stopped reading there. The images are being accessed by the browser by sending a request to the Cara URL. None of it is being cached by HF's CDN or the likes.
The URL itself existing and being shared is not a problem, it is what let's us access the image that was uploaded (and compressed) by Cara to the platform publicly for viewing.
Stopped reading there. The images are being accessed by the browser by sending a request to the Cara URL. None of it is being cached by HF's CDN or the likes.
Sorry! I made a misclick, discussions are open!
I would highly recommend you read the rest if you would like to understand better and properly discuss this, I cannot communicate my point properly if you are unwilling to read (sorry if this sounds rude, for I have no intent to be such).
This is the screenshot of the provided link incase for whatever reason it is unviewable.
snakegoy69420, the huggingface repository only has URLs, not images, the size of the dataset is a bit over a gigabyte, in your "proof" the total size of all the images is 12TB
This is the screenshot of the provided link incase for whatever reason it is unviewable.
snakegoy69420, the huggingface repository only has URLs, not images, the size of the dataset is a bit over a gigabyte, in your "proof" the total size of all the images is 12TB
This is true! And I am not denying the fact that this dataset hosts URLs to data, not direct data.
However, this dataset is actively encouraging further copyright infringement by mass downloading and training, and although you may disagree with that, it is absolutely undeniable that this event has harassed and caused a lot of mental and physical issues to peoples of Cara.
This means to say - the dataset is not innocent because it holds URLs, you may be confusing copyright of what here because it is a URL. We are not protecting the URL from being visible, we are (attempting to) protect the works directly linked to via the URL.
However, this dataset is actively encouraging further copyright infringement by mass downloading and training
Copyright infringement as in downloading a publicly accessible image to one's computer for private viewing? Lol.
it is absolutely undeniable that this event has harassed and caused a lot of mental and physical issues to peoples of Cara.
Get mental health counseling then.
However, this dataset is actively encouraging further copyright infringement by mass downloading and training
Copyright infringement as in downloading a publicly accessible image to one's computer for private viewing? Lol.
it is absolutely undeniable that this event has harassed and caused a lot of mental and physical issues to peoples of Cara.
Get mental health counseling then.
Copyright Infringement has a definition of (from Wikipedia):
Copyright infringement (sometimes referred to as piracy) is the use of works protected by copyright without permission for usage where such permission is required, thereby infringing certain exclusive rights granted to the copyright holder, such as the right to reproduce, distribute, display or perform the protected work, or to produce derivative works. The copyright holder is usually the work's creator, or a publisher or other business to whom copyright has been assigned. Copyright holders routinely invoke legal and technological measures to prevent and penalize copyright infringement.
This is the type we are referring to, and are in no mistake in assuming it!
As for your second comment, it shows you are merely here to just say something, while I am not directly attacked, many other people are, and if you don't see a problem in that, I am completely in my right to stop communicating with you, in interest of avoiding your subjectively disgusting behavior.
I hope you have a good day.
for usage where such permission is required
And nobody cares about your subjective point of view. Please do stop communicating, since you are only embarrassing yourself. If you weren't some underage knee-jerking fool, you would've filed a tangible copyright infringement notice in the form of a DMCA email, rather than making a bogus "report" under discussions.