The idea is to have a bot that can take part in debates against human speakers. It would listen to their speech, transcribe it, formulate a response in a debate format and use text-to-speech to give its own speech for the audience to hear.
What I've learned so far:
- How to apply my knowledge of Python
- How to use Faster Whisper AI model for transcription
- Why virtual environments are useful
What I expect to learn:
- How to make API calls to OpenAI
- How to use Text-to-Speech AI models like Kokoro
Current progress:
- It's possible to record human speech in english and have it transcribed to text files. Also displays it in console
- Started work on adding API calls to OpenAI to get a response
Next up:
- Improve code design
- Improve the API calls
- Add Text-to-Speech using Kokoro
- Add an interface to control the behaviour and actions of the program
- Improve the speed of transcription
-To be added...