3 ms·
I doubt an in-house model will be able to get enough training data. I doubt public domain data plus their own personal copyrighted stuff is remotely sufficient.
by WalterBright 4y ago
I doubt an in-house model will be able to get enough training data. I doubt public domain data plus their own personal copyrighted stuff is remotely sufficient.
- krapp 4y agoI'm sure Disney could get their hands on enough Disney owned material to train a Disney AI. It would also be a reason to keep artists and staff on hand to create training material for new properties, so maybe not as many creatives would be out on the streets. Plus I'm assuming IP owners would have deals with stock image libraries and companies like Adobe to make it actually legal for companies to use.