This should come as no surprise to anyone who has been paying attention to Google, but it’s a reminder to get off of their services if you’re concerned about your data privacy. A Slashdot commenter wondered whether this also applies to the contents of emails on gmail. I’m going to assume it does.
All llms are made with scraped data, all image generative services as well.
If data were to become restricted and not public property the moment it’s posted online, then only Adobe and Shutterstock could legally make an image generation service.
Don’t fall into the game of trying to restrict public data, it will only hurt us, the little guys, in the long run.
If posted to a public forum like this one, non-commercial usage is obviously required. On the biggest (i.e., for-profit) social platforms, commercial use is also required (if I’m understanding those terms correctly, at least).
There is stuff publicly available online that has restrictions on commercial use. Because I’d rather see how open source AI generation goes, I’d prefer to have non-commercial efforts get that sweet data without giving it to the major tech companies.