Ben Congdon expands AI voice workflows using STT and TTS

The author argues that speech is an underutilized medium for interacting with AI, offering practical workflows for input, output, and real-time interaction. Using tools like local STT models and text-to-speech generators, the author demonstrates how voice can streamline tasks such as coding agent prompts, ambient meeting transcriptions, daily customized audio briefings, and conversational flashcard reviews. Open article →