Meta stands accused of pirating over 80 terabytes of copyrighted books from platforms like LibGen, Anna's Archive, and Z-library to train its AI models. In response to the ensuing lawsuit, Meta reportedly argued that "individual books themselves are, essentially, worthless," and that a single book contributes "less than 0.06%" to an AI model's performance. However, this seemingly small percentage equates to significant value. Consider an AI model valued at $1 billion; a 0.06% contribution from a single book represents $600,000 in value. To put this in perspective, Meta's valuation stands at approximately $1 trillion, while Claude is valued at $61 billion, and OpenAI at $300 billion. Adding to the complexity, the US Copyright Office has indicated that copying even a minor portion of a copyrighted work for the purpose of training AI models does not constitute "fair use."
Download the medial app to read full posts, comements and news.