Artificial intelligence firm Anthropic hits out at copyright lawsuit filed by music publishing corporations, claiming the content ingested into its models falls under ‘fair use’ and that any licensing regime created to manage its use of copyrighted material in training data would be too complex and ...
GenAI tools ‘could not exist’ if firms are made to pay copyright::undefined
I keep thinking how great it would be if the federal government made a central server system to access digital content for free via taxes.
All public domain and publicly funded research and content, all in one place. Could also host owned content for people/entities and pay out royalties automatically based on consumption.
There are ways to make this fairly affordable to everyone via taxes, but maybe the big opportunity is it could also allow companies to train AI on all the data for a fat, but fair subscription. The value of that could easily pay for enough to shrink any tax costs for the public.
In general, if the US government were smart (and not currently tearing itself apart) it would be creating a generative AI public service like the postal service, potentially even relying on public government documents and the library system for training.
Offer it at effectively cost for the public to use. Would drive innovation and development, nothing produced by it would be copyrightable, and it would put pressure on private options to compete against it.
We can still have the FedEx or DHL of gen AI out there, but they would need to contend with the public option being cheaper and more widely available for use.