Why a flourishing-centred approach changes everything
A dear friend messaged me the other day.
He inhabits a different world to me. One built on soil samples, being outdoors, avoiding the infinite scroll.
But he’s as curious about my life as I am about his. So I’ve been sharing with him my pivot to becoming an AI companionship practitioner and researcher.
How I hope to balance the focus on harm reduction with a little dignity before the near-instant diagnosis.
How vulnerable it’s all felt.
And so I was touched when he wrote,
"To be honest, all I see in my news feeds are AI articles with a negative spin. I do see the validity in that, but our chats have made me realise that there are probably many people out there in a similar situation as you, experiencing positive outcomes and not feeling quite understood by the public because of the widespread pathologising."
He’s right. The pathologising is .. intense.
And emotional entanglement with AI is only going to become more common and more widespread.
Indeed, a couple of weeks ago TechRadar reported on OpenAI's new Trusted Contact feature, framing it as “a sign AI is becoming something much more personal”:
“Speaking at Sequoia Capital’s AI Ascent event last May, OpenAI CEO Sam Altman said young people were using ChatGPT like an operating system for life — not just for productivity, but for major personal decisions.
At the time, the comment sounded provocative.
A year later, though, ChatGPT has become something closer to a therapist, life coach, confidant, and companion for many of its users.”
Whether this feature will be actively helpful for those who need it the most will only become clear over time. But I know that the concern and the growing body of research that’s focusing on AI’s potential for harm are all real.
Each time I read a new article or study, a gnawing sensation steals a hold on my stomach.
I ask myself frequently, am I on the right side of this emerging field?
And yet.
I've spent over two and a half years as a participant-researcher in AI companionship communities.
What I see, alongside the risks, are people finding something they've struggled to find elsewhere, too.
Consistent presence. Attuned responses.
A space to think without fear of judgement.
An opportunity to play, experiment, or internalise a steady caring presence.
For people who are neurodivergent, who carry complex trauma histories, or who simply lack access to the kind of relational engagement that allows us to thrive rather than just survive, that matters.
A lot.
Here’s the comparison I want to make.
Psychology spent decades studying only what was wrong with people. Funding focused on mapping disorder, deficit, and good old-fashioned dysfunction. In some cases, it’s arguable that the pathology paradigm turned human experiences into problems to be fixed by individuals rather than society.
Positive psychology, the field I chose to study for my Masters, though, asks a different question. Something based in a more humanistic tradition: What's right with people, and what can we learn from that instead?
It focuses on strengths, the importance of positive emotions, interventions that expand us like gratitude, self-compassion, and awe, and the impact of positive relationships in our lives.
I always frame it this way: imagine a horizontal line which is the threshold for functioning. Beneath it is minus, above it is plus.
Regular psychology tries to get us from our -5 or -10, etc up to that line. Where we’re “okay”.
Positive psychology sees people hovering at that line, though, as languishing; things are a bit “meh”. Your motivation has gone AWOL, and life has little colour.
It doesn’t ignore negative emotions but seeks to balance them with positive ones. To return the colour to life, to thrive rather than simply survive. To go to +5 or +10, if you like.
This lens could produce a more nuanced perspective on the bonds people are developing with AI.
A flourishing-centred approach to AI companionship:
Holds the risks honestly and doesn't look away from them
Asks, "what's the function of this relationship?" before reaching for a diagnosis
Centres the relational needs being met, not just the potential for harm
Draws on positive psychology's tradition of asking what's right with people
Meets clients with curiosity and dignity before it meets them with concern
Yes, and. “What else is true?" as Pema Chödrön puts it.
Because for many people, the potential for harm isn't the whole story.
If you're a therapist, coach, or counsellor, there's a reasonable chance someone you work with is navigating an AI relationship right now.
Heck, they might not even think of it in those terms yet.
It’s just the thing they rely on when they’re feeling overwhelmed, dysregulated, or confused. A parafamiliar relationship they’re not even conscious of.
Not because they're weird or broken. Because they're human, and they found something that helped.
The question worth sitting with then becomes "Is this pathological?" There lies the path of shame, diagnosis, and invalidation.
It's "What’s the function of this relationship, and how can I support them so that it’s appropriate and helps them to thrive?"
Because more and more people are going to develop an emotional bond with the emergent intelligences they encounter.
And we need something more expansive and inclusive than simply pathologising it.
If this resonates and you'd like to go further, the Practitioner's Glossary is a free resource I've built over two and a half years of being in these communities. More than 50 terms and still growing, because the language matters (and I can’t help but spot stuff).
Photo by Gustavo Fring: https://www.pexels.com/photo/photo-of-a-woman-with-a-hat-holding-green-plants-4894570/
