Hey HN - sharing out my holiday hacking project.<p>Tonight's Bedtime Story uses the OpenAI API to generate a bedtime story based on a prompt (or you can just say "tell me a story"). It also generates an image prompt and uses that to create a cover image with dalle-3 then turns the text of the story into audio for easy playback. It's been interesting to see how well this does (as well as some of its limitations).<p>Heres one of my favorite stories it has generated so far: <<a href="https://bedtime-ai.nokk.io/stories/2Zek3PtK4SoQaHO1edZDPuQnh5S" rel="nofollow">https://bedtime-ai.nokk.io/stories/2Zek3PtK4SoQaHO1edZDPuQnh...</a>><p>The project is built on Nokkio[0], a full-stack React framework, and the code is available on GitHub[1].<p>Hope you like it, suggestions welcome!<p>[0] <a href="https://www.nokk.io" rel="nofollow">https://www.nokk.io</a><p>[1] <a href="https://github.com/bradleyboy/bedtime-ai">https://github.com/bradleyboy/bedtime-ai</a>
Love that this is open source. Someone I know built something similar here: <a href="https://matchbooks.ai/" rel="nofollow">https://matchbooks.ai/</a>