You can now prompt ChatGPT with pictures and voice commands
OpenAI’s excellent Whisper model does a lot of the speech-to-text work, and the company is rolling out a new text-to-speech model it says can generate “human-like audio from just text and a few seconds of…