The problem is simply that voice is a 1-D (linear) interface whereas computer screen is 2-D. So when I interact with the screen it can show me all my possible

The problem is simply that voice is a 1-D (linear) interface whereas computer screen is 2-D. So when I interact with the screen it can show me all my possible choices at the same time at each situation. I click a menu-item and a sub-menu pops up. With a voice-assistant I would have to wait for the assistant to list all possible choices but if they are many that takes a long time. You can only hear one word at a time. Combine that with inaccuracies of voice-recognition and it is clear that a computer screen + mouse is a much better interface. Hey it's the same difference as with Radio vs. TV. A computer screen-interface might be usefully augmented with voice recognition. But working with voice alone is like working blind. Gee who could've thought voice is not the killer-tech of 3rd millennium.