Comment on NVIDIA: Copyrighted Books Are Just Statistical Correlations to Our AI Models * TorrentFreak
LallyLuckFarm@beehaw.org 4 months agoIirc the leaked internal docs and reporting showed that nvidia was downloading 81 years worth of video per day. If the pattern holds, and we “throw” the books at them, it probably looks a lot like one of those building-sized front end loaders dropping the library of Alexandria on those executives.
As an alternative, I would be fine if nvidia were no longer allowed to hold copyright or trademarks since the company believes that those don’t matter, required to share all internal documents (including any and all planning), and required to open source all their code until the company implodes