Using YouAI, I created an AI chat bot and trained it using an open source collection of books on meditation. I gave the bot the role of "guided meditation script writer." Using the open source material it was trained on, its job is to receive a user-provided topic and then write a guided meditation script based on that. For example, if given the topic “letting go of regret”, it would write a guided meditation where it helps the meditator focus on letting go of their regrets.
I then gave it some topics and took my scripts.
Next, I took the completed scripts to Eleven Labs, a cutting edge text to speech generator. Using the voice “Emily”, I fed in the scripts and downloaded the resulting audio files.
With all the AI magic complete, I took the files to GarageBand for editing. I mixed the audio narration with GB’s built in audio library assets, as well as some freebies from Pixabay to create the final audio files.
I also created artwork (not generated by AI) for the playlist and each track.
You can find the results hosted on SoundCloud, and I've embedded an audio player below. The result isn’t perfect. First of all, I’m not a professional audio producer, and I’m not trained to guide others in meditation. I was feeling very uneasy about these two facts as I started on all this. But then I tried out what I made, and it worked really well for me. How fantastic to be able to produce a guided meditation audio file that basically addresses just the things I want to focus on, and also does it in a style that I prefer. The audio mixing at the end takes a good amount of time though, so this is far from automatic. The voices aren’t perfect either. It’s pretty clear that all this is changing quickly though, so I expect everything I’m doing here will get antiquated soon enough, the quality will improve, and more of it will be automatic. For better or worse, as with all things AI-related.