Benj Edwards turned a voice memo about hauling moldy trash from his mother’s basement into an AI-produced funk song. He still plays it in the car. His children dislike his generated music, and he doubts anyone else would want to hear it. Its value comes from the memory attached to it.
He calls this personal media: songs, games and other entertainment made chiefly for their maker. Suno let him turn roughly 50 old recordings into hundreds of arrangements. Claude Code helps him build games he enjoys playing, including Violent Checkers and Card Miner. He describes the appeal as a hobby, alongside model kits or aquariums.
The economic argument is straightforward. Expensive production needs enough customers to pay for it. Cheaper production lets a project serve one person’s tastes without earning back a band’s wages or a studio budget. Edwards connects this to Robin Sloan’s home-cooked software: making something for a known, tiny audience changes what is worth making.
He also revisits his earlier prediction that abundant generated entertainment would destroy its commercial value. Shared cultural events still give people reasons to watch the same film. His revised expectation is coexistence: a few enormous hits above a growing layer of small, strange, personal works.
Private enjoyment leaves the labor question open. Edwards acknowledges that cheaper production could devastate creative jobs. Another possibility is rising output expectations: people work harder because their tools can do more. He raises the need for worker protections and ends without claiming an ethical resolution.
That leaves a narrower, persuasive claim: something can matter deeply to its maker while having almost no market value. His experience supports that claim. It cannot tell us how audiences will divide their attention, whose work trained the tools, or who will capture the gains from cheaper production.