• TehBamski@lemmy.world
      link
      fedilink
      arrow-up
      3
      ·
      4 months ago

      Most likely AI. You can train generative AI to learn how to speak like just about anyone. They just need hours of recordings to do it well. How much? No idea. But a good amount for sure.

      • Tyoda@lemm.ee
        link
        fedilink
        arrow-up
        5
        ·
        4 months ago

        But a good amount for sure.

        Nope. Nowadays the required audio source is measured in seconds.