Am I the only one who doesn't want a phone that arbitrarily changes behavior in opaque and uncontrollable ways? Machine learning is fine for hard tasks like voice recognition, but it really shouldn't be used for things that would otherwise be a few checkboxes. I know the trend nowadays is to assume that users are irredeemably stupid, but this? Seriously? You can keep this update. I don't want it.
I keep thinking of that famous quote about debugging but in the context of AI: "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it."
I think the writing is on the wall for the way Google has been going on about AI first solutions, but Google has failed to acknowledge it. There was a lot of backlash from Youtube monetization, Youtube kids, censored search results of certain topics. All Google can do is claim they will try to improve "The Algorithm", but at the end of the day they are distancing themselves from having control of their own outputs.
AI allows nice to have features we couldn't have otherwise, but is it ok for all stake holders to have such a feature with the condition that it won't work around 1% of the time, or for 1% of users?