✌️Dual-mode speech and canvas

My take on how speech to speech UX might evolve. An assistant would speak a 'talk track' with supplementary content in a dynamic canvas.