Any AI company wanting to stay competitive in building gen AI models will need to be training on more data than before - images, music, video, etc. Which data they are training on and the Copyright implications of that is a big question. Another question is what happens now that much of that is AI-generated, does it degrade future models?
nice article. Which companies do you think will still need to train on large datasets of image, video and audio in 2025?
Any AI company wanting to stay competitive in building gen AI models will need to be training on more data than before - images, music, video, etc. Which data they are training on and the Copyright implications of that is a big question. Another question is what happens now that much of that is AI-generated, does it degrade future models?