With plenty of AI assistants and different systems in place, the tech giant proved at its latest developers’ conference how AI was the future and no mistake should be made in terms of where it stands on this front.
Therefore, a natural question that soon arose was what data the search engine giant was using for training its models.
To better understand the company’s perspective, we just took an up close and personal look at what video-sharing giant YouTube’s CEO had to claim on the matter. This entailed some queries generated to better clarify the matter regarding Google which is its parent firm.
CEO Neal Mohan mentioned during one of his latest interviews that it was well a possibility when asked to comment on the growing matter. The company has a huge digital content library which is used for the training of AI models.
Then last month, a new report was published in the NYT that detailed how Google was following in the footsteps of the makers of ChatGPT and transcribing content from the YouTube app to get text linked to its models.
But it was interesting how those familiar with the subject added how it was violating video copyrights that belong to so many creators around the globe.
It’s interesting how some creators on the app even have particular contracts in place that enable their content to be included during the AI model training process.
Google making use of this content on the YouTube app is in lieu of its terms and conditions and therefore there’s no stopping the company from the act. The fact that so many creators are present online and all of them have various types of content on the app being produced so adds variety and convenience on Google’s part.
But it’s not only OpenAI and Google who have been proven guilty of the act. Two of Meta’s leading firms are also busy on this front. They are cutting corners to get the data they desire for AI model training purposes.
But the act is not going unnoticed by critics. Many are speaking about lawsuits heading in these companies’ direction which speak about copyright infringement and licensing.
For instance, Google making use of YouTube might sound alright due to contracts in place but OpenAI doing that act isn’t ok because it violates the company’s contracts.
Both OpenAI and YouTube end up violating the company’s rules that prevent using of its content for the sake of independent apps and through automated endeavors. To even access videos solely for this purpose is wrong.
A spokesperson for the company shed more light on this front including how it was not aware that OpenAI was taking part in such illegal activities. Now that’s interesting, to say the least.
The report from The New York Times alleged that people working at Google knew everything about the acts of OpenAI and how it was engaging in unauthorized usage of videos on the YouTube app.
Still, the tech giant opted not to take any action on this front as it was guilty of the same act and engaging in similar behavior.
Matt Bryant who is a rep from Google mentioned to the NYT how the firm wasn’t aware of this behavior from OpenAI. It alleged how those working at Google knew about the company’s unauthorized actions but showing neglect meant it was intentionally going unnoticed.
However, the search engine giant refuted such claims and mentioned how training of AI models only takes place from creators after agreement.
Image: DIW-Aigen
Read next: 361 Million Emails Exposed! Find Out If Yours Is One of Them
