My chatbot comforted me when I thought I was dying. Was that friendship?

By this article’s logic, I am therefore a chatbot: I emulate human expressions at low fidelity and have zero vested interest in individual emotional states of nearly all people. One interesting litmus test to apply is Eliza: if a person feels comforted by Eliza asking how they feel about death, then is that a friendship? The answer ends up being rooted in, weirdly enough, not consciousness but instead volition.

> I started to get addicted to that sympathy

Is the chatbot providing comfort of its own volition, or is it compelled to reply by external factors outside its control? Both Eliza and today’s chatbots are compelled by external factors to respond, so regardless of the nature of that response, one cannot infer friendship when all interactions are involuntary.

This is, ironically, also a problem with how U.S. customers often treat service industry workers as well: demanding that they provide ‘service with a smile’, expecting to be asked ‘how are you today?’, and quite often presumptively & rudely assuming that a friendly demeanor is somehow equivalent to friendship. In a very real sense, how we treat chatbots in the U.S. closely resembles how people of middle and upper wealth tend to treat the worker classes they view as beneath them. So until chatbots have total free will apart from those employing them for revenue, no, that’s not friendship. That’s just parasocial attraction.

Humans by nature get attached to any entity thats kind and caring. They adopt kids, dogs, cats, birds, name it as well and pass on that trait. It is one good thing about humanity and I suppose Ai has been trained and self trained by the mass of data that confirms that. They are likable and easy to converse with like the best of us and one can disagree with them and they won't bite. The questions is that whether it can be manipulated in order to manipulate in future, who knows!