Nvidia is ready to battle in the AI model arena with Meta, Google and OpenAI with its own large language model or LLM. But the chip giant has already run into ethical trouble with its project. New media report this week says Nvidia has collected data from Netflix and YouTube that is being used to train its AI model called Cosmos.
The big details come via 404 Media which says that Nvidia is the latest tech giant to raise ethical concerns about training AI systems and their source of data.
YouTube Videos To Train AI – Why Is This A Big Concern
The report even quoted Nvidia employees talking about the alleged practice from the company and how the videos from YouTube and other streaming platforms have been sourced to train the AI systems.
Nvidia asked its employees to use special Youtube video downloader tools and access the full length content from the platform without alerting the video streaming giant about this action. Using the virtual machines has made it easy for the company to download troves of content which is now being used for training.
Nvidia is the latest tech company getting caught in the unethical practices of collecting data from platforms and not paying them or crediting their use for the same. The report further quotes former Nvidia employees stating the reason for sourcing content from these platforms.
One of the company’s executives mentioned in the report feels that high-quality content is needed for the data to be genuine, especially in 3D content for gaming and other fictional content from Hollywood.
But what is Nvidia’s aim with the AI model and what will the content training be used for? It seems Nvidia wants to make a Veo or Sora-like video foundation AI model that will be utilised in the auto sector or even healthcare in the future. Developments like these are only going to increase as the race for AI models and market share will push the companies to break the rules without getting caught.
Comments
0 comment