v21’s avatarv21’s Twitter Archive—№ 70,818

        1. is it fair use to train ML systems on data you don't have the rights to? what counts as "public data"? @natfriedman/1409914420579344385
      1. …in reply to @v21
        what if your system sometimes reproduces source material verbatim? @natfriedman/1409921379542081541?s=19
    1. …in reply to @v21
  1. …in reply to @v21