Maybe because you haven’t seen an AI first designed ‘anything’. I doubt they really have a sense of what it is either, but if they actually did take what is incorrectly, but popularly, phrased as ‘AI’ and built a personal communication platform from it, I think it would be different enough that you saying ‘it’s not worth it’ before having any sense of what it is, is premature in the most literal sense.
Mostly it’s because, information wise, it’s almost nearly “free” to take a design and duplicate it…bilateral symmetry is natures version of copy/paste.
With that in mind, it’s likely that non-‘bilaterally symmetrical’ organisms relatively regularly spontaneously develop it due to random mutation. Just like we often randomly find people with extra fingers or only one set of organs, over millions of generations, bilateral symmetry will naturally just happen. The difference being, extra fingers or ‘more than two’ organs rarely offer any evolutionary advantage, especially in already complex forms.
Millions of years ago, however, very simple organisms suddenly having two brain lobes, two eyes, twice as many fins, two gills, etc….for free (informationally) and at only a relatively higher cost energy-wise could have found itself at a distinct advantage. If you can both run from predators and towards food twice as fast, and the energy cost isn’t twice as much, you’re suddenly the two legged guy at the ass kicking contest in a parade full of one legged people.
There’s a tension and maybe responsiveness to skin and muscles that is uncanny when missing. Not sure many here could 100% recognize that very early on at the point of death, but at some point there is a wariness/unnatural look to the skin. Between that and our assumed ability to pick up on a complete lack of movement/breathing/pallor makes it reasonably certain that there is a “something” we recognize as missing, even if it’s hard to describe perfectly.
If it’s just a verbal interface to a smartphone it’s going to be a waste of time. There are a lot of people who do feel comfortable blabbering their thoughts out loud regardless of their surroundings, but that seems to have a big overlap with people wanting attention.
If it’s truly ‘AI’, it should be able to incorporate what truly works for people, whether that means speech to text for outbound messages, summarizing long emails for inbound, gestures, haptics, anticipating time based tasks, to making up meal plans when it recognizes you’re adding random items to your shopping list and looking up a dozen recipes, and figuring out what alarms and alerts actual get your attention for things you actually treat as important vs the ones you mark as important and then snooze a dozen times. If it actually starts with AI, it might recognize what alert you need to see on your computer and what notifications it can wait to show when your on the toilet….that future is awesome and scary and will probably make some billionaires before it wipes out humanity or turns us into infants crying to have our diapers changed as it takes over everything else.