Howdy.
Last year I began developing a simple text to speech reader for my Son who has Dyslexia.
Nothing too fancy. Just a simple google extension built using Open Ai API to power the Ai speech.
The extension got abandoned mainly because the TTS kind of sucked. But, this did give me an idea for a larger application.
So my dream would be something like this - Turn text into audio with a simple paste or drag-and-drop. Your files are stored securely using decentralised protocols. Codex. And then once stored have the ability to share privately through encrypted peer-to-peer channels.
The original extension was named LOGASTELLUS.
I’m a designer with no real deep tech experience so would love to hear some thought on this, feasibility, direction etc
Thanks 
2 Likes
(as I already said to you) I love the idea - especially since I already tried to prototype something like this to be able to consume long Forum threads in a form of audio:).
So far the biggest blocker for me was to find a decent TTS model that would run locally - most that actually work well are cloud based (e.g. Elevenlabs) - if anyone has any suggestions, I am all ears:-)
I am talking to Dria (https://dria.co/) about adding TTS models to their AI network and they promised, so let’s see “when”
Otherwise, if we can find anything decent and reasonably runnable (that was another caveat that anything that looked promising, I was not able to run because of dependencies and random errors:D), I would love to try to integrate it with Podex - i.e. apart from uploading audio/video directly, you could also upload text and get the audio generated - I think that would be a neat fun feature.
Thanks for posting this!
PS: I also love the name:)
2 Likes
