The voice of the computer, part 2

Here’s Matt Webb col­lecting his thoughts on voice inter­faces—rich and circumspect, with tons of links, tons to think about.

My posi­tion remains that voice is a daz­zling demo that mostly cur­dles in prac­tical use.

What about voice unvoiced, though? I can for sure imagine some kind of del­i­cate sub­vo­cal­iza­tion pickup becoming pretty irresistible … yet that only con­jures a world of people silently wig­gling their neck mus­cles at their phones, which I don’t think is much of an improve­ment over the world we already have.

As I write this, I am realizing: I wish more designers and dreamers would con­sider the aggre­gate social effects of the inter­faces they imagine. Maybe I even wish they would imagine that shared scene FIRST, then work back­wards from there.

To be clear, I agree that the foun­da­tional metaphors and modal­i­ties of com­puting are about to change — and they were overdue for change anyway. The framing and ambi­tion of Telepath (new to me, dis­cov­ered in Matt’s links) seems exactly right. I just don’t think voice gets us to that next thing. But/and, as I said before, it’s pos­sible that Google and OpenAI are seeing people absolutely swarm to their super­fluent voice modes, in which case … I might be wrong 🥸