A core hypothesis of a lot of AI/automation feature design seems to be that it will supercharge more user agency.
-
A core hypothesis of a lot of AI/automation feature design seems to be that it will supercharge more user agency. We'll "remove x task" and therefore free you to do y task.
This is a testable hypothesis that just isn't tested that often imo! Like, I have been experiencing this REMOVAL of agency so much of the time. I was just using a note taking tool and had some text filler automation shim in SO seamlessly that it disrupted and confused me for a second ("did I type that??"). Loss of agency
-
@grimalkina and increasingly you have two competing sources of such “help” - a given application and the OS of your device (iOS or Android offering “help” autocorrecting your inputs. As a user I’m not always sure who to blame for such “help” like Apple (which my phone just helpfully capitalized) or auto inserting spaces in compound words - even when I really do want the compound word not two words with a space as the full word is relevant to the domain I’m writing about
Loss of agency indeed
-
@Rycaut yeah totally!! There are so many layers of "decisions" being made invisible and opaque to users here -- this is a point that's made often in tech communities I feel but extraordinarily out of reach for many people because so many forces are conspiring to make "invisible choices" the ultimate goal of design
-
@grimalkina indeed and I recall some research (from years ago but I suspect the numbers haven’t shifted much) that showed that a very large percentage (80-90+%) of users (across all platforms of software) use all software with the defaults as installed (whether from the developer or however their enterprise set defaults). Both the OS itself and each app they use.
While software devs (and the “power” users they interact with most) are apt to fully customize each app and tweak options