Let it decentralised. Ai will learn directly from source and pay for it without any centralised authority     See Less

Everything seems unfair in the beginning. The road is long.     See Less

he will get that law passed, why? cause China AI would advance since they dont care about west copyright, so it becomes a matter of "national security"     See Less

If you want to compete with China then you need AI to train on copyrighted material, cause China doesn't give a damn about copyright. I can see why this is fair.     See Less

Restricting AI from accessing copyrighted materials is a big misunderstanding and an even bigger mistake. Arguing against allowing AI models to train on copyrighted materials, suggesting it     See More s plagiarism or infringement. Misunderstands entirely the essential nature of how AI learns and generates creative outputs. An to answer your question yes it will matter in 18 months as it might impede AI access to most recent knowledge and information.    See Less

US copyright law was written for the expressly to the advantage of the public not the creator, (i.e. more available work to the public) so the logic follows then that AI training is merely a     See More ce of the logic not a violation of it. As such any copyright holder would have no claim against AI training.    See Less

In other words America will be left in chinas dust what i want to know is how much of deepseek are they going to use in their next model     See Less

We may have exhausted all publicly available data, but obviously not *all* available data by any means. If public models have been taken as far as they can with public data, th     See More eaves a largely untapped market of organizations who will need help training and maintaining these public models (and custom private models) on their own internal, private data. Perhaps that’s where folks like myself come in who work for consulting firms and systems integrators but OpenAI could partner or pivot to extract some value from that revenue stream. Maybe not quite a multi-billion dollar a year business they were hoping for but depending on how they position themselves and their offerings, potentially so.    See Less

So I can get sued for sampling a fart from 1970 but…     See Less

First. China already caught up with the US in AI. They just released within last month 4 new engines that work better and faster than GPT. But at least they are making the code public and gi     See More to the public    See Less