2 ms·
Then we should put the same restrictions on OpenAI as we do as humans. So if a website has a paywall, OpenAI must pay to view the content. However, since this d
by jessenaser 3y ago
Then we should put the same restrictions on OpenAI as we do as humans. So if a website has a paywall, OpenAI must pay to view the content. However, since this data is being given to an AI that is not an individual, there probably be different licensing so copyright holders can extract value from their works in the final model in some way. Maybe some payment before, and some after.
Even non paywalled content should receive compensation of that work is copyrighted. OpenAI should not be able to profit off the work of others at a mass scale in this fashion.
- ang_cire 3y agoFirst, we should not give AI models the same rights as humans, because they're not humans. We should place far MORE restrictions on AI models. Second, we should force OpenAI to cut deals with each content creator whose content they want to make use of.
- jessenaser 3y ago1. Correct 2. Yes. Since even if GPT-4 is producing transformative content from "itself", a new medium of profit was created (training AI models) which have not been done before. (Example what I mean:) Even though a library is free, you are not expected to go into the library, and read 10,000 books within a few hours, and put the books back on the shelf, and walk out like it is fine. Humans can only listen, watch, and read only a finite amount of content in their lifetime, GPT-4 can read at a scale that eventually be all humans on earth reading at the same time. Streaming services for music only exist because you cannot scrape all the music in the catalog. You will listen to few songs to pay the artists and record label more percentage in comparison to the whole catalog you wont listen to.