Ironically in early 2023 a lot of websites went out of their way to block Common Crawl. Unsurprisingly that shifted scraping toward individual actors whereas the previous solution in research was to download CC dumps and process them.
I don't think it'll blow over as easily because the PS6 is going to be as expensive as a PC[0], at $1,000 or more. The usual things that inhibit consumer action are hassle, expense, and inertia, and Sony is dishing out all of that to its customers already.
"America or not" is a red herring. Palantir is the specific problem.
The kind of people running these companies don't have true allegiance to anything but their own objectives. If necessary, they'd move all operations to Europe or East Asia in a month, and you'd have "Palantir 2" under a different name with no better ethics or privacy.
Increased emphasis has to be on running things domestically with on-premises hardware. As long as the vendor is elsewhere and not subject to oversight, the risks remain.
I've addressed why these schemes are questionable before. [0]
"This doesn't work because AI companies don't accept licenses. Either the training is already infringing (in which case this doesn't matter much) or it's fair use/permitted by law (in which case this also doesn't matter much)."
Those are the exceptions that prove the rule. It's very difficult for US or Western companies in general to do business in China without opaque restrictions, corruptions, and share ownership hoops. If the US is playing games, then it's closer to kids playing soccer on weekends; China is already in the pro leagues.
https://undeleted.ronsor.com
https://gpu.st
email contact: "ronsor" + chr(64) + "ronsor.com"
Operator of AS211233. Try me: `ping6 2a0a:6040:8e00::1`.