Where does voice actually fit in how people use AI?

Something we keep noticing: even when speaking a prompt would be faster, we still type it.

Voice feels great for quick asks like "set a timer" or "what's the weather" and awkward for anything that needs revising mid-sentence, or reading back a complex answer.

There is a real case that LLMs make voice replies less scripted, closer to actual conversation than what came before. And there is also the case that text has been the default input for computers for 50 years and there are reasons for that. Precision, editability, silence in public, references you can scroll back to. Voice does not give you those.

We do not have a strong read on where this lands in 5 years. Maybe it is not winner-take-all.

When do you actually reach for voice with AI, and when do you type?