- What is OpenAI's level of copyright now?
- How is it going to be more "copyrighted" in the future?
- How does this affect competitors differently in the future vs. the copyright that OpenAI has now?
Limited. They’re hoping to change that. It’s no secret that open-source models are the long-run competition to the likes of OpenAI.
Nobody does. It’s being litigated.
They want it legislated. Model weights being proprietary by statute would close off the threat from “consumer-grade hardware” with “exaflop-scale.”
Then why did you say "Limited"? Surely YOU must have meant something by it when you said it. What did YOU mean?
I don't think you're saying that you are repeating something someone else said, and you didn't think they knew what they meant by it, and you also don't know what you/they meant. Correct me if I'm wrong, but I'm assuming you had/have a meaning in mind. If you were just repeating something someone else said who didn't know what they meant by it, then please correct me and let me know -- because that's what "nobody knows what it means" implies, but I feel like you knew what you meant so I'm failing to connect something here.
> It’s being litigated.
I'm not able to find any ongoing suits involving OpenAI asserting copyright over anything. Can you point me to one? I only see some where OpenAI is trying to weaken any existing copyright protections, to their benefit. I must be missing something.
I'm also unable to find any lobbyist / think-tank / press release talking points on establishing copyright protections for model weights.
Where did you see this ongoing litigation?
[1] https://www.bereskinparr.com/doc/chatgpt-ip-strategy
[2] https://hbr.org/2023/04/generative-ai-has-an-intellectual-pr...