In this video above, you’ll see how to go about API integrations to create a voice-based AI chat by connecting multiple APIs together. If you’re interested in a detailed walkthrough of the process, be sure to subscribe to my Patreon page to see the full tutorial.
Youtube: Best bubble developer
Patreon: Bestbubbledev
Twitter: @bestbubbledev
LinkedIn: Gio Kakhiani
Yes, it’s very easy to use Chrome with webkitSpeechRecognition() and SpeechSynthesisUtterance()
Made a record button (as a POC) that is injected into ChatGTP which does STT => TTS (pilots the interface) via a chrome extension (may relase it soon)
I created my own GPT-3 assistant where I can send a text message and it responds back. Pretty cool when out in the field an no WiFi. For voice, just use the built-in voice-to-text in iPhone.
Very Cool idea! 
combining old and new tech to solve a real problem! Could be useful for older people also!