Companies using YouTube videos to train models face backlash

From NDTV Profit.: 2024-07-20 10:37:57

Several leading AI companies, including Nvidia, Anthropic, and Apple, are under fire for using subtitles from over 173,000 YouTube videos to train their models, sparking outrage from affected content creators. The dataset was compiled by EleutherAI, a non-profit AI lab. Video creators invest time and effort into their content, only to have it repurposed without permission for AI development.

Representative Jennifer Wexton of Virginia’s 10th Congressional District has regained her voice through AI technology after being diagnosed with Progressive Supranuclear Palsy. She now uses an AI-generated voice model based on her previous speeches, offering hope for others facing similar challenges. This advancement goes beyond traditional text-to-speech solutions.

OpenAI has introduced the GPT-4o mini, a more compact and cost-effective version of its GPT-4o model. Priced at ¢15 per 1 million input tokens and ¢60 per 1 million output tokens, it offers a 60% cost reduction compared to its predecessor, GPT-3.5 Turbo. The model boasts an 82% score on the MMLU benchmark and supports up to 16K output tokens per request.

As AI technology progresses, there are new possibilities and advancements on the horizon. Stay updated on the latest developments in the field through newsletters like “Beyond Tomorrow,” offering a weekly roundup of AI news and breakthroughs.



Read more at NDTV Profit.: Companies Train Models On YouTube Videos, OpenAIs New Mini Model — Weekly AI Roundup