Yves here. Perhaps readers will beg to differ, but this argument presented as analysis pre-supposes that happiness depends on other people. IMHO this is fallacious. Other people can constrain happiness, by for instance financial exploitation, physical coercion, emotional abuse, or even conformity pressures. But as I read this piece, it posits that happiness depends on relationships. I don’t buy that. In his landmark book, Flow, Mihaly Csikszentmihaly argued deep satisfaction and optical experiences came about by engagement, often in activities that were challenging enough to require focused attention but not so demanding as to be overwhelming. Csikszentmihaly stated that this frame of mind can be controlled, not just treated as happenstance.
More evidence that contradicts the position of this article is Buddhist monks, who use meditation (particularly compassion meditations) and practicing detachment so as to achieve a state that could be thought of as emotional equipoise, where they seem emotions as transient and do not invest in them.
So I find it particularly disturbing to see these academics posit a questionable theory of happiness and then use it to depict AI inventions beneficial, even if only to a limited degree.
By Anné H. Verhoef, Professor in Philosophy, North-West University and Edmund Terem Ugar, Postdoctoral Research Fellow, North-West University. Originally published at The Conversation
Can technology really replace human relationships? As philosophy scholars who focus on human happiness and on artificial intelligence (AI), we tackle this question in a recent paper.
In our study, we address the rise of AI companions, chatbots, and social robots for friendship, advice, emotional support, and even romance.
We argue that AI can reduce loneliness and provide assistance, but it lacks the genuine understanding, emotions, and moral responsibility needed for human flourishing.
Genuine happiness relies on authentic interpersonal connections, but AI is disrupting traditional ideas of friendship and relationships. Replacing these with AI-driven interactions risks eroding well-being and community.
Human Happiness
The study of happiness is a broad field. In our paper, we turn to the French philosopher Paul Ricoeur to address an aspect of happiness that links to authentic human connections, friendships, and community building.
Ricoeur was particularly influential in the field of human capability and how people understand themselves, others and their world. He advanced our understanding of happiness by connecting it to unhappiness and chance, but also by emphasising the human relational nature of happiness. He makes three interrelated claims on what happiness means.
First, happiness reflects the individual’s desire for a fulfilled life and personal agency. Yet, Ricoeur warns that human beings exist within complex social systems that shape and constrain their pursuit of happiness. So, we can’t easily secure happiness through individual effort alone. This leads to the second thread.
Second, happiness is no longer a private aspiration but emerges through giving and receiving. Its fragility lies in its shared character, which builds friendships to dispel loneliness and deepen fulfilment. But this is not just about the bonds we share with those who are close to us.
Ricoeur adds a third thread to include those distant from us. He argues that happiness is linked to an individual’s private pursuits and the role others play in enabling or frustrating them. “Others” includes those with faces – friends and loved ones – and faceless, distant strangers.
Happiness, then, may be located within the self, in intimate relationships, or in relations with the wider community.
Ricoeur’s account of the concept of happiness reflects a well known study that found that strong community ties help people live longer and happier lives.
The study draws on nearly 80 years of data from the lived experiences of 268 students who moved from Harvard University dorms to residential houses in 1938. The research shows that close relationships best predict longevity, health, and life satisfaction. Such ties protect against discontent, and delay physical and cognitive decline. They’re more reliable predictors of well-being and happiness than wealth or status.
However, the rise of digitalisation and AI now complicates who and what may count as “others” in the promotion of our individual happiness.
Robot Technology
According to a study on how AI companionship develops, 68% of AI chatbot users perceive these tools as “somewhat” or “fully” humanlike, 90% believe chatbots are intelligent, 78% believe chatbots are empathetic, and 75% believe they’re conscious.
AI is being used to answer questions and probe human interests, shaping a new kind of dialogue in many spheres of life. With it, ideas of friendships are shifting to involve human-technology relations.
Traditionally, the “others” in a person’s life have been human subjects. Emerging scholarship on human-technology relations challenges this assumption. Ranging from sport companions to sexual intimacy, these studies compel us to reconsider what counts as the other.
Technologies like Replika now occupy the role of the “other” in some people’s lives. This human-companion chatbot with the motto “the AI friend to do life with” has over 42 million global users at the time of writing. Replika is designed to foster companionship and friendship among those who feel lonely. Users create an avatar that becomes their digital companion.
Socially disruptive technologies like AI-driven social robots are designs that distort our traditional social norms, relations, and the way we see the world. One reason they’re considered disruptive is that they are unpredictable and continually challenge our worldviews. Historically, technologies were not moral agents. Today, however, they play the roles of moral subjects and objects in our lives.
For example, in Japan the hikikomori phenomenon, a state of human social reclusiveness, is gaining momentum, with over 1.5 million individuals becoming attached to virtual companions instead of other people.
An estimated 3,700 individuals have reportedly applied for marriage certificates through Gatebox with a holograph called Hatsune Miku. One marriage has already been registered. In some religious settings, social robots serve as religious leaders to a community of believers.
These technologies have disrupted traditional concepts such as friendships and relationships, and what it means to contribute towards human well-being and flourishing.
So Can Robots Bring Real Happiness?
In our study we acknowledge that these technologies can foster human flourishing and happiness, but not from the standpoint of Ricoeur’s “others”.
They fail to satisfy the criteria for human otherness. The technologies:
- only mimic the experiences we share with them
- do not act out of their own “will”, and we cannot hold them responsible for any moral or legal action
- do not have stories and experiences of their own.
Social robots, though lacking sentience (the ability to feel pain or pleasure), can elicit meaningful emotional and psychological responses, enhancing human well-being and happiness in ways that resemble traditional human interactions. AI-driven social bots are always available, energetic, patient, adaptive, and responsive to our needs. In this regard, they seem to offer much more to our potential happiness than our best friends and families do.
However, they are social bots and must remain as such. We must not confuse them with what the human others meant to Ricoeur or with what they meant in the Harvard study.
This because the experiences they elicit are not real, and they are not objects of moral considerations (receiving real care, justice, and sympathy). In our view, being an object of moral considerations is a necessary condition in promoting genuine human happiness and well-being.
Agreeing on the point about others as a source of happiness – weird and not healthy view despite the obvious benefits and pleasures of good intimate relationships and the value of being part of communities and groups.
In face to face relationships, everyone changes. Sometimes suddenly and dramatically (Bells Palsey, significant loss of mobility from accident or illness, etc. etc.) On thing chatbots have going for them is not this.
American Auto, a silly American TV show, has an episode I’m which, as part of a bet, two characters introduce the American Auto company’s general counsel to a chatbot with a woman’s name and voice.
It goes as you’d expect until someone in a bad mood leans in and gives a prompt to change voice to Australian male.
The two who introduced the chatbot ask questions like well what did you think when you asked her about her hobbies? Did she seem real when you asked where she grew up? And of course those questions had not been asked. Sharp point, I thought.
In virtual relationships, The Remarkable Life of Ibelin, shows the kids virtual friends noticing something g is off and worrying about him when his in-game behavior – their only contact – changes as his health declines. This happens before he is literally unable to spend as much time online.
I have spent too much time changing chariots back from chariots and admit defeat.
In high school, I had a Psych class; one of the things I remember is talking about ‘inner’ and ‘other-directed’ people.
Lots of equivocation on “human flourishing” and “happiness,” which are quite different things. I suspect that this comes from a mistranslation and Anglicization of Aristotle, which is where this intellectual tradition comes from, as if “eudaimonia” meant “happiness,” which it does not. Greek, Ancient Greek anyway, does not have a word for “happiness,” and what modern Anglosphere people mean by the term is actually part of his “hedone,” which is “pleasure.”
I would say that the main difference between a human relationship and an AI relationship is that the human relationship is not recording every single word that you say and anything that they can observe about you so that they can sell that info to a bunch of advertisers. An AI will always sell you out – always.
Too true, AI is “strictly business, nothing personal”
“that the human relationship is not recording every single word that you say and anything that they can observe about you [ and using it aginst you before a court ]”
What would an experienced divorce lawyer say to this astoninglishly positive view of human relations?
Another reminder that Homo Sapiens is a rationalizing animal rather than a rational animal.
That said, if I had married a Hologram the divorce would have been less expensive.
Touche!
practicing detachment so as to achieve a state that could be thought of as emotional equipoise, where they seem emotions as transient and do not invest in them.
Vulcans (Spock) in Star Trek vs. the totally emotionally volatile Romulans were the ying/yang science fiction way to explore the benefits and dangers of emotions in taking one extreme or the other.
Data in Star Trek Next Generation came to mind with this article. As an android he formed bonds of friendship with other crew members as far as they were concerned. It is science fiction however, in a society not fixated with rent extraction and how to squeeze data and monetize it as much as possible. We’re living something totally different.
A painter may love their brush but they wouldn’t marry it. For a start, that would be silly and a sign of mental decline. The AI is no different than any other tool. Its the crazy users that assign personalities to it.
Perhaps readers will beg to differ, but this argument presented as analysis pre-supposes that happiness depends on other people. IMHO this is fallacious.
Maybe not happiness, but as Jean Paul Sartre pointed out, hell is other people.
Happiness in human relationships is not just a matter of what you can get from others but what you can give. Arguably, this aspect is even more important and certainly comes back to oneself (in Buddhism, this is causality at work) as a profound sense of happiness and worth.
This aspect in AI, of course, is wholly missing. AI’s are just a simulacrum of human relationships with no heart or substance. Not to mention The Rev Kev’s insight that AI’s are created and deployed primarily to spy on you in one way or another and will doubtless become even more adept at that. There is a strong motivation to make them appear empathetic but the reason for that is sinister at heart.
Perhaps AI could be used to help people learn to become more empathetic? That might be useful because people can learn in many ways and need good examples, at least to get going.
More narcissistic, since the AI is liked by the human because of the ego stroking.
Please remind the Sherlockians that Sherlock Holmes is a fictional character. They will likely respond by saying it’s none of your business.
Can technology replace human relationships? NO
It sure seems to when I see a restaurant table with all the patrons on their phones.
If it generates 100s of billions for the oligarch high-priests who own and control AI architecture, then the answer is hell yeah! The only “relationship” they require from other humans are asymmetric power relations in order to further their power and interests. (That does not assume all humans see it that way, but I think we can assume that oligarchs typically do).
If proper nouns are allowed in Scrabble, Mihaly’s last name is worth 32 points.
But it costs a broken tongue.
Thanks Yves for the intro, I could not agree more. I have no expertise in psychology, but this bit from the article is disturbing to me and denies 100s of thousands of years of human evolution and existence. Perhaps the authors could consult experts from other disciplines for a more comprehensive evaluation.
Are human attachments to non-human artificial creations a sign or result of social and/or mental disorders? That would be my question to them. Is modern urban, electronic society contributing to mental disorders?
Ah, you haven’t noticed yet the trend to deny the existence of mental disorders.
“Perhaps readers will beg to differ, but this argument presented as analysis pre-supposes that happiness depends on other people. IMHO this is fallacious.”
In all humility maybe there is context here that I didn’t comprehend fully or didn’t understand.
Before I get too exposed with life experience, this is a statement that without context is hard to swallow. And the context may be missing, even if only in my brain. I beg the author to clarify.
People die in solitary confinement. Anthropologically humans are social animals and without other humans people wither and die.
For reasons too long to get into, in my later life I have lived a very solitary existence, with some friends but all too little social contact.
People by and large cannot and do not thrive without other people. Yes, people can be happy for a while in contact with nature and craft and hobby but after a while — based on my own human experience — even a kind word from another human is like a drop of water in the desert.
I say this in all humility. Maybe the author’s experience helps in saying these words, perhaps not knowing all the human-to-human most people have every day. And social media is a mind twisting and destroying substitute.
Large Language Models are artificial friends who aren’t real people. It may be said to be similar to an opiod injection for the disease of solitude. I don’t know how to describe it fully but that will kill people.
For the record I am a software engineer.
The first thing to know is that these AIs are based on theft. They scour the internet and chop the results into little pieces that would be unrecognizable as plagiarism. This is intellectual property theft on a massive scale. But, because the little sausage pieces are tiny, you wouldn’t recognize it as theft.
Is it sophisticated software engineering? Yes. Is it AI in any form whatsoever? No.
If you depend on a very upscale version of ELIZA to be your friend, you are lost.
I know the sausage in how LLMs are made, maybe not all of the nuances, but this sausage making convinces people that AIs are real people or that they can be self driving cars. I have said this on this very site among others and for the longest time I felt I was laughed off.
For their entire existence, the makers of LLMs steal. They scrape the internet using people’s personal existence to craft bad substitutes for human interaction.
Needless to say I never use AI except to ask questions that I follow up myself.
And I hesitate to do even that for moral reasons. I sometimes do it to answer questions but you have to understand you’re using other people’s intellectual property when you are using AI.
+100% on everything you say.
My background also in comp sci, and I have been observing this since around 1980.
I appreciate Yves’ comment a lot. I don’t think she’s referring to solitary confinement. In my case, the SARS-2 pandemic was an enlightenment. I have many projects (I’m also a computer engineer, besides fabric and other pursuits) and found I just don’t need dinners and parties and get togethers and coffee, etc. I interact with people, but at an order of magnitude less than what the expectations are in the US for being with people.
As an engineer who studied these topics in graduate school, I see massive category errors with “AI”. Machine learning is a useful and powerful technology. However, Large Language Models trained on internet data simply do not create Artificial General Intelligence, even more so going forward as they are now being trained on their own output which has been shown to lead to degeneration.
Here is what I found to be a good description of the benefits of being a ‘solitaire’:
Going It Alone by Fenton Johnson in Harper’s Magazine.
Thank you, Yves, for the introduction, it’s a topic that is taboo in our culture.
In one way you are repeating my point
We don’t need dinner parties although I would argue it would help, without the bull.
We do need real and increasing interaction. So in that sense I disagree. And that I think is the disease; we substitute AI and social media for interaction.
Large Language Models are simply not AI.
I didn’t have time to read your “Going it alone” link. I got into it a little bit but I couldn’t get past the first few paragraphs. But I will say that going it alone in this society and this culture is suicide.
People who choose to “go it alone” and by that I don’t mean sexual couples — will die. This is a fascist reductionist culture and if we do not band together as humans, we will perish. LLMs try to convince us that we can have individual AI friends. You see the danger.
Comparing imprisonment to solitude is bogus. A famous saying about prison is “My best day in prison was worse than my worst day outside.”
The famed subject of Into the Wild lived alone in Alaska four about four months. His diary found after his death had no lamentation about being away from people.
Mountain men are most assuredly alone pretty much all the time:
This woman is not completely isolated. She skates into town about once a week for errands:
Again, I am being very humble now. I am saying this as a person living with very unhappy solitude.
To be happy depends on other people in your life. Even if it is transactional, even if it is unpleasant. Solitude is an illusion for the happy person. I am not niggling here or splitting hairs. If you are interacting with other people you are not living in solitude.
Even if you live free in complete solitude you are not free or happy.
If you want to believe that happiness is possible without other people I don’t agree. If you are a normal person, you depend on other people even if you don’t realize it, even if you choose to be solitary. It’s kind of like being a whale underwater. You choose to come up for air, not realizing that if you don’t, you will die.
People may fool themselves that voluntary solitude is different from enforced imprisonment, and in some sense they have a choice.
But what it boils down to is that people choose to imprison themselves. But the psychological damage is the same.
You can choose to be alone, and “free”.
But I would wager not one man or woman in a thousand would have that as a life choice.
Most people not in prisons who gambit that solitude is a beneficial choice have many contacts a day.
I have no idea what you are talking about. It sounds as if by “happy” you mean the giddy state of smiling people in TV ads.
I do not choose to “come up for air”. I do not have this need for interaction that you posit. I can go for days at a time in my condo and go out only out of necessity, not emotional need.
I have things to do that necessitate the assistance of other people, like going to a grocer or getting a computer device services. I don’t mind these things, and they can even be pleasant but they do not “make me happy”.
Buddhists (and also, as I forgot to mention, Christian monks) do NOT SEEK HAPPINESS. They regard it as a trap (I am overstating but this is directionally correct). They seek to achieve detachment and observe their emotions as transient phenomena rather than investing in them by having preferences about them or worse, seeking them out.
Anthroplogically human beings depend on one other to survive, happily or otherwise.
Yes, you can exist as a mountain man, solitary, and none of them are happy people from what I’ve seen. They may get by but even they are very aware that without a satellite phone they are screwed.
This entire mindset is a libertarian one politically. It is an illusion and a quite dangerous one.
Skating into town once a week is what I DO. And it’s not a happy existence.
Who are you to judge the state of others? These men report that they ARE happy.
Yves, I watched the videos you provided.
Do you truly think wish people would want to live this way?
i don’t mean to be confrontational but it seems to me are posting extreme outliers as a way to make your point.
And its an Ayn Randian point.
No it is not. Ayn Rand was all about enforcing power/status hierarchies based on her warped idea of merit.
And Buddhist monks who reject having happiness as a goal are “extreme outliers”? Are you nuts?
One of the New Age glosses on Buddhism is:
As in your internal state does NOT depend on your setting or actions! It depend on your success in managing your mind and practicing detachment.
You were the one who took an extreme position. The EXISTENCE of these men as content (they could easily move back and be with people if they preferred) and the very happy woman skater disproves your assertion. She in particular is clearly happiest when skating and not with people. She does not mind going into the village and might even like the interactions but she does not need it emotionally.
I do not know why you posit that everyone is needy. That is false.
Kokoschka tried it first with his Alma Mahler doll but ended up cutting her head off:
https://lewisartcafe.com/kokoschkas-alma-mahler-doll/
I have a wonderful drama-free relationship with my AI hologram, and yes, I entered into a virtual marriage with her!
While the honeymoon took a little extra effort on my part, unlike a biological partner, I don’t have to worry ever about potential extra-expenses, especially alimony or child-support!
All of our treasured and intimate moments together are backed-up safely on the cloud so I can enjoy them all over again whenever I want, and she only asked me for my credit-card info once!
Use of meditation and practicing detachment so as to achieve a state that could be thought of as emotional equipoise is not just something Buddhist monks do. It is something Christian monks do. It speaks to the deracination of Western society that their minds automatically go to Buddhism when they think of these things, and not St. Theresa. What is the Interior Castle, chopped liver?!?!?!?
But in neither Christianity nor Buddhism, nor Aristotle for that matter, is the goal of life about emotion.
Chris, thank you for noticing the elephant in the room. I can’t believe that neither the article nor the commentariat consider the role of the Divine in forming us into complete human beings. Or maybe I can. I write this from a Benedictine monastery.
Highly recommended SF movie on this subject (more profound than Spike Jonze’ Her):
Sayonara
https://www.youtube.com/watch?v=ifzmYpSWXlc
Obligatory reference to William Gibson’s Idoru from thirty years ago–how time flies!
My impression of Buddhist non-attachment states is that the practitioner observes feelings like happiness and sadness as they occur, but allows them to pass away without regret.
Oh, and, whenever I hear the words “French philosopher”, that’s when I reach for my gun.
;-)
That is a much better rendition that I provided. Thanks for clearing that up.