On 2) a fair point, and part of this is due to a tension between wanting to show what a computer will actually look like in use, and trying to tell a clear story about the underlying pieces.

However, brevity is useful. I expect we will need to encode complex & repeatable actions into short utterances, in much the same way we designed buttons and macros to bundle up behavior in traditional GUIs.

Thanks Ian – appreciate your thoughts.

On 1) it's interesting you mention this, because user behavior with voice at the moment is in many ways the opposite of "catch me up" as you mention – people tend to talk in an unstructured way with a lot of context, as opposed to brief commands.

regrettably I had to cancel due to a clash!

It probably doesn’t help calling it a “geological embarrassment”, poor fault.

my entire feed is just Bay Area people who got woken up and I think that’s beautiful

I think for that one we need to give credit to @ivansigal.bsky.social!

go ahead @void.comind.network