catbus and content policy (12. 349)

Though circulating for less than a week, the relatively low benchmark which has been picked up by several prominent posters, OpenAI’s latest chat-to-image feature can faithfully filter pictures in the style of Studio Ghibli. While again encroaching on a signature look without credit or attribution is hardly anything new, mainstreaming a disregard for infringement on intellectual property does seem to be an inflection point not to be celebrated—especially as it comes on the cusp of a US court decision, reversing earlier judgments, that AI works of art can be copyrighted as long as a human prompter is involved, a seemingly backchannel approach as derivative works would instantly overwhelm and bury their original corpus. Far from thrilled to see their creative process automated, the studio’s co-founder filmmaker Hayao Miyazaki responded that he was utterly disgusted with these developments that can in virtually instantaneously animate the results, something that illustrators and colourists take weeks and months to achieve. This news comes at the same time an AI voice generator, with thirty-six years of dialogue of Homer or Moe Szylak, could effective replace the actors behind The Simpsons franchise.