you sorta missed the point. A system like this learns and adapts. It's an assistant, similar to Siri but more personalized, more tuned in with your actual needs and not busy selling your secrets for extra $$$.
When the TV remote was first becoming a thing my grandfather would rant that it made no sense to have such a thing because all you had to do was get up and go turn the knobs yourself.
Progess is trying to make people's lives easier, more comfortable and more fulfilling. It's not like there is a conspiracy to replace humans. It's just that if a task can be automated it should be.
A perfect example. Alice uses SMS as her prefered method of communication but Bob prefers email.
Alice can be cognizant of this fact, or her phone can. If Alice has to be aware of this, then she must send Bob an email. If she sends him a text he might get pissed.
If her phone can be aware of it then all her phone needs to do is know bobs phone and ask it "hey alice has a message for bob, how should I send it"?
Alice doesn't need to worry about the message carrier and can focus on the message content, same with Bob.
RE: Personal Preference Bot Nets and The Quantification of Intention