Cryptocurrency
AI Dataset Controversy in the Crypto World
Wednesday. July 17 at 9:00 AM
1 min. readEleutherAI, a non-profit AI research group, reportedly violated YouTube's terms of service by scraping subtitles from YouTube videos to create a dataset called the Pile. This dataset, consisting of subtitles from over 173,000 YouTube videos across 48,000 channels, has been used by tech giants like Anthropic, Salesforce, Apple, Nvidia, Bloomberg, and Databricks for AI training. The dataset also includes content from crypto channels like Coinbase, Cointelegraph, and Bitcoin Magazine. The controversy extends to AI copyright disputes, with lawsuits involving companies such as Anthropic, Meta, GitHub, Nvidia, and Google. The article also mentions the rise in crypto lobbying, with Coinbase leading the way. It emphasizes the importance of conducting thorough research before engaging in cryptocurrency activities, as they are considered high-risk ventures. The crypto world continues to evolve, with new developments and challenges emerging regularly.