Conversation

Notices

1. Embed this notice
   camwilson (camwilson@mastodon.social)'s status on Tuesday, 12-Aug-2025 13:48:01 JST camwilson

   much of the discussion re: AI & copyright has been about whether training is fair use.

   that's a legal debate that I'm less interested in than the question about what happens to the sources of information — news, culture etc — if a lot of its value can be extracted by AI without much compensation

   • Embed this notice
     camwilson (camwilson@mastodon.social)'s status on Tuesday, 12-Aug-2025 13:48:01 JST camwilson

     in reply to

     If you think AI training is the same as someone reading and reciting, that's fine fine. but our legal system wasn't set up for a world in which there's like, what, 5-10? major AI models which become a major market for information because they are the middleman between consumers and info producers

     Steve's Place repeated this.
   • Embed this notice
     camwilson (camwilson@mastodon.social)'s status on Tuesday, 12-Aug-2025 13:48:01 JST camwilson

     in reply to

     debating copyright law misses the point that the new paradigm of information distribution is extremely centralised. their size and the amount of information ingested was never imagined when thinking about fair use. we need to new ways to pay for info that's clearly valuable to the AI companies