The Last Barrier to Thinking Out Loud

Dictation makes you better at using AI because talking produces richer, more specific input than typing. But there's a catch sitting right in the middle of that insight: most people won't do it consistently because they feel awkward talking out loud in shared spaces. The habit that unlocks better AI output is gated by social friction.

Silent dictation — mouthing words and having them captured from facial micromovements — removes that gate entirely. It preserves everything that makes dictation powerful: the natural specificity, the uncompressed thinking, the conversational flow that models respond to so well. It just strips away the part where other people can hear you. If this technology matures, the argument for dictation stops being "you should try this at home" and becomes "this is just how you interact with AI, everywhere, all the time."