AI and copyright

In addition to a model card (seriously, they are the only major AI lab without one), it would great for xAI to report anything about the grand experiment of letting Grok be an arbiter of truth on this site. How accurate is it? How helpful? The effect is important to understand.

This is literally the book trusted by Stanford, MIT, and Google to teach AI pic.x.com/MiJ7ZDqZdR

Chris Bryant today all but admitted why the government refuses to accept legislation that makes AI companies reveal their training data: They are going to offer creators transparency legislation *later*, in return for upending copyright law & letting AI companies train on pic.x.com/c2Z2Xyb0xC

The UK government said it didn’t have the money to implement a law that would require AI companies to disclose their training data Yet it has enough to spend £187 million teaching “AI skills” 🤔