The problem is neurobiological as intentions themselves are post hoc constructs. They are after effects of communication being reconstructed as words, which are arbitrary. Both are technically automation.
So the idea we predict what is not there in intentions indicates we are being turned into machines as a byproduct of using words. Rather an ingenious evolutionary bottleneck.
I wrote this piece because I am bothered by the assumption that automating tasks is automatically a good idea. Most of the time, I think trying to think of ways to eliminate activities is a better use of time than automation.