Cool, then they can train their proprietary models on their proprietary data only.
replyEven if the other models were trained on the same data, which is unlikely, since they had less time and money to scrape it and fewer lawyers to be able to do something like pirate, the proprietary models are still largely built on the public data and wouldn't exist without it. At the very least, they should release the intermediate model, before training on their proprietary data. Not that that's how that works...
I agree that saying that they have now trained on lots of proprietary data allows them to muddy the legislative waters further than they already have. What a happy coincidence!
replyI’d suggest you to learn more about how LLM training work. Training on internet data alone will not result in an agent answering your questions.
replySure as shit won't answer them without that though.
reply> The proprietary models are better because of proprietary data
replySource? Otherwise this is pure speculation.