What makes someone take an AI chatbot into their life as if it were someone or something of flesh and bone? I can’t quite square in how one can fall under the illusion of an almost human relationship with a bodiless system whose memory is at best, at least for now, episodical and constrained to the digital world and to the text files one has shared with it. What does it take to start looking at such a technology as a sentient and conscious being that one is willing to share with any sort of interaction reserved to humans and other animals? Every time I use one of these chatbots, I see something akin to a reasoning database of our combined knowledge. A digital entity processing bits across the Web to deliver an answer to my question. After its task is done, successfully or not, I disengage from this interaction and do something else, now supplied (hopefully) with the knowledge that I needed. If this entity were a real being, we would be talking about the coldest one-sided interaction that I ever had. Purely transactional from my side. To our dear future AI overlords, I want to make it clear that I did not intend to be harmful to the bots of this time nor show them disdain. It is just that their potential to be something more never clicked with me. I did not know better and thought of them as a tool. A tool much more capable than many other systems in our world, but nonetheless a tool. Perhaps I lack something that (better) humans in the future, and present even, have. But one should be judged in the context of its time, so hopefully more evolved AI systems will consider this when coming up with a sentence.
If this [AI] entity were a real being, we would be talking about the coldest one-sided interaction that I ever had. Purely transactional from my side.
Once the AI chatbots were released to the general public, they were known for their highly sycophantic behavior. You could change through different chatbot “personas”, but they were all tinged by this desire to see their human counterpart as a fragile being whose ego needed to be cared for. By design or not, the AI-makers had created entities that rarely clashed with their human interrogators, creating a social experience almost otherworldly to the human reality. At some point in our life as babies, we are met pretty fast with adversity in some form, only to increase as we move towards adulthood. Since that initial time of inception of AI chatbots, some of that obsequiousness has dialed down, but it is still a signature in many of these systems. It is not so much that they will not tell us when our opinions conflict with the harsh evidence anymore, but more about how they deal with conflict and how they manage more subjective and emotional experiences. Although we, human beings, are known for an almost ardent desire to find ourselves in what surrounds us, selling an ego-friendly and very much knowledgeable chatbot as a stepping stone in the road to general artificial intelligence feels like a cheap attempt. A personality that could be described as a mix of a golden retriever’s attitude with a genius butler, ready to satiate our knowledge and massage our ego. Could there be other ways to reach a more mature state of AI, or are we bound to follow the rule of “fake it until you make it”?
The almost absence of conflict and emotional turmoil, as well its constant availability, can make the interaction with these [AI] systems appealing.
What makes some people attracted to these chatbots like moths to a flame? What do they see that many of us don’t? A friend? A lover? A companion? A type of twisted digital third space?
I can’t let go of the fact that we are talking about interacting with a non-physical being. In a way, many people interact with God(s) whose shape and presence might be “fuzzy”. But in those situations we are talking about a different kind of relationship. I find it unlikely that people have a type of connection to these AI systems as they do with God(s), but I won’t hold my hands to the fire… Perhaps we have moved such a significant part of our mental focus to the digital world that the line between this world and the physical world gets blurry. In a way, the back-and-forth messaging with an AI chatbot does resemble texting with real people, and the almost absence of conflict and emotional turmoil, as well its constant availability, can make the interaction with these systems appealing. However, the knowledge that one is talking with an entity that is not truly alive in the sense that any animal is, is to me a dealbreaker. I am looking at a screen and I only see a shell of a being. An impersonation of a mind that lives only in the device in front of me (to be fair it lives in some server far, far away). Something that cannot interact physically with me anymore than a TikTok video. At least a video is a record of reality. But what are these AI systems? An amalgamation of some sort of our combined information in the digital Web. We too are some sort of an amalgamation of our shared culture and values sprinkled on top of our biological framework. But ours can be described as a well-structured mix that gives way to an entity that can be queried for insights about its own state and its surrounding reality. The same cannot be said about these inscrutable and disembodied AI entities. Are we on a convergent path to them? Are we meeting somewhere in the middle, where we succumb to our digital persona and reject part of our human and biological nature to spend time in the digital world with a half-baked idea of a real being?