[ad_1]
A YouTube content creator filed a lawsuit in the U.S. District Court for the Northern District of California on Friday, alleging Open AI Use millions of YouTube video transcriptions to train their generative AI models without notifying or compensating the video owners.
Millett’s lawyer (YouTuber’s name) claims Open AI Transcribing videos from Millette and other creators to train driving models ChatGPTand other generative AI tools and products. In doing so, they point out Open AI “Benefit significantly” from the work of content creators.
Mira: Find out which iconic phones appeared in The Matrix and why they’re so memorable
this Require Imposing means that the behavior violates copyright and Terms of Service Youtube Use of Platform Content for stand-alone applications is prohibited.
“Most of the material in the training dataset Open AI From copied works Open AI No consent, no credit, no compensation,” the lawsuit states.
The lawyers are seeking to take the case to court and obtain more than $5 million in damages to compensate users and content creators for their losses. Youtube.
Mira: Elon Musk resumes lawsuit against OpenAI: He accuses of treason and fraud
Why does this demand arise?
In a survey published in April by The New York Times, the media announced that by the end of 2021 Open AI There is not enough data to continue training his model AI, ChatGPT.
With this in mind, they looked for a way to solve this problem. So, the engineers created Whisper, an AI model that can transcribe audio. YouTube Video This will allow them to continue training their Artificial Intelligence System.
The problem is that this activity infringes upon YoutubeLindsay Holder, Spokesperson Open AI Announced to The Verge “The company selects ‘unique’ datasets for each of its models to ‘help it understand the world’ and maintain its global research competitiveness.”
It also added that it uses “many sources, including public data and non-public data correlation”, and that they are investigating the possibility of generating their own synthetic data.
This activity not only carried out Open AIas the same Times report announced Google Also collected Transcript of Youtube Train your own Artificial Intelligence ModelHowever, Google spokesman Matt Bryan said they have used “some Youtubeaccording to our YouTube Creators”.
at present, Open AI No comments have been made on this matter yet Class Action Proposed by Millett.
GDA / El Universal / Mexico
[ad_2]
Source link
