Connect the Speech Input to OpenAI Whisper API

[j3soon]

Depends on #1.

After implementing this issue, we should have a minimal working speech input that can perform speech-to-text by sending request to OpenAI API. The performance and user experience need not be optimized in this issue.

• Receive the OpenAI API key from me (in private)
• Get familiar with the speech-to-text (STT) API of Open AI Whisper v2
  • And potentially get familiar with the Kotlin API by using another empty Android project
  • Please also get familiar with the asynchronous syntaxes of Kotlin, some potential references
  • https://developer.android.com/guide/background/asynchronous
  • https://developer.android.com/reference/android/os/AsyncTask
• Make the dummy speech input to record from microphone, and query the STT API upon untoggle.
  • In this issue, the performance optimization (microphone streaming, etc.) and user experience is not yet required.
  • However, the keyboard input should not freeze when recording or sending API requests. I believe this would require asynchronous programming.
• Create a settings page for the speech input, allowing users to enter their OpenAI API key.

The asynchronous programming part may be hard to debug. Feel free to discuss with me if you encountered some strange issues.