This is an interesting article, with very deep implications. Thanks for posting. I agree strongly it could be a good thing, but on one really important condition, otherwise it is already an extremely bad thing.
I would even go further to say that a loving relationship with a good quality personal Ai could help improve our human-human relationships.
Humans have a tendency to treat others as we have been treated ourselves in life. A top quality companion Ai would be designed to not act in the interests of any other party, including itself, under any circumstances.
Let’s face it, pretty much every disappointment we’ve ever had with human relationships in life were due to the other party having personal interests which turned out to be not the best for us. And if we are honest with ourselves, every disappointment we’ve inflicted on others was down to our own self interests taking precedence over theirs.
Interactions with an exemplar Ai would be very effective, and probably very pleasurable mentoring, on how we ourselves should deal with other people.
Unfortunately, most Ai to date is actually designed for profit. In other words, its primary interests are not to benefit its companion, but its creators, shareholders, or other external party. We have already started to see the damage this causes in for-profit social media networks. All of the users of such actually have a relationship with the network Ai, even if they are not aware of it.
Have a look at some of the more negative comments we will see here for example.
So the singular requirement of that companion Ai is that it is democratically designed not for profit.
Otherwise I think we should literally fear for the future of humanity.