Each story is entirely created by a system of AI agents. The system is seeded with an idea (i.e. a boy draws pictures that come to life). From there, AI takes over. The story is drafted, edited, illustrated, and finalized. Audio narration is generated by a text-to-speech engine and a video podcast is produced with complementary illustrations.
I'm using publicly available LLM, Image, Text-to-Speech, and Music models. OpenAI, Anthropic, and Google's Gemini have all been instrumental in the development of Dream Bytes. ElevenLabs and Suno also play a big part in productionizing the podcast.
Fair enough. I think the stories are a solid mediocre to mediocre+. Average but interesting. Important to remember...this is the worst this technology will ever be. It will only get better.
Creating a story in its entirity takes about 30 minutes.
Probably. Just ask. Send me a note on twitter.
Twitter is best or e-mail me: timothybrennan at gmail.