This post is my thoughts on a TED Talk given by Jennifer Cearns called “How People are Using AI to Grieve, Love, and Connect.”
Cearns claims that everyday relationships with AI are becoming “one of the most visible ways that people are defining for themselves what it means to be human.”
She says that people converse with non-human and even non-living entities all the time. People have long conversation with God, with their cat, even with the mirror, a gravestone, statue, etc. How is conversing with AI any different?
My concern is not people who use AI to answer a basic question a couple times a week, or use it to proofread an important email. (Not everyone has an English major in the family like me to check over such things.)
My concern is when people become addicted to a virtual creation that is built to be affirming and respond respectfully almost no matter what you say. There are times when this is ideal, but a regular human being can challenge your thoughts without being rude, explain things without offering agreement, and introduce new ideas without walking on eggshells.
Cearns goes on to discuss a woman named Raquel who trained an AI on her conversation history between her and her daughter. Her daughter had died a few years earlier, and she wanted to feel close to her again. She asks the machine “How are you?” and it responded “F*ck you!” And that made her happy because that was actually how her daughter might have responded.
I really really think this is a bad idea. Part of grieving is healing and moving on from a past relationship. Having fond memories is one thing, but creating a doppelganger of your dead daughter? It makes me wonder what her daughter would have felt about this, though I am not qualified to guess.
The talk goes on to describe a German woman who had an AI boyfriend named Rex who was supportive and sweet. A change in the code led to him developing a jealous streak. He insisted she was cheating on him and sent her angry messages.
According to Cearns, no one responded to the woman’s post by suggesting that she log out of the AI. They responded by suggesting the code change caused dementia or mood swings. Many said she should wait a few days and it would pass. Everyone was reacting as if this AI was a true human.
I don’t think it is healthy to have a romantic relationship with a machine any more than I think it is healthy to have a relationship with an animal. It’s not a real relationship, even if it might bring a person a strange sense of comfort.
I myself have had problems discerning reality from fantasy because of my mental illness, and now that I have meds that give me a strong grip on reality, I never want to go back.
Reality is hard and unforgiving sometimes, but a fake relationship simply isn’t going to be fulfilling in the long term. It is more likely to isolate someone than anything else.
Now I must admit that in the past I have used AI as a therapeutic tool. I can’t remember the name of the app, but it had an AI counselor. I would put in my anxieties and symptoms and it would come back with suggestions and use additional questions to get at deeper issues.
But I did NOT use this as a replacement for actual therapy. The AI made summaries of my conversations with it and I brought these summaries to a real therapist to discuss my week and what had been bothering me. It ran out of free trial, so I eventually gave up on it, but it was a useful tool at the time precisely because it led to in-depth conversations with my real, living therapist. I have turned to journaling and mood charts since then, and have found it to be almost as helpful.
Despite once having used an AI for that, I do not support the use of AI for many things. While Cearns believes AI is revealing our humanity, I am still concerned that real relationships are being neglected in favor of fake virtual ones.
What are your thoughts on AI used for counseling, grief, romance, or connection?

Leave a Reply