OpenAI claimed it’s “impossible” to build good AI models without using copyrighted data. An “ethically created” large language model and a giant AI dataset of public domain text suggest otherwise.
Here’s Proof You Can Train an AI Model Without Slurping Copyrighted Content
