What Happens When We Begin to Relate to AI? When AI Changes: Part 2
In Part 1 the interaction that appears to be relational yet AI has no capacity for relationship, were explored. Here we'll continue that exploration.
What happens when we begin to relate to AI becomes even more significant when the technology itself changes.
An AI interaction can feel remarkably consistent. But the technology underneath it is not fixed. Models may be updated. Safety settings may change. Response styles may alter. Memory may work differently. Features may disappear.
An interaction that has become familiar may suddenly feel unfamiliar.
Now imagine that the interaction has mattered emotionally.
Perhaps it helped reduce shame. Perhaps you began listening to yourself differently. Perhaps the familiar responsiveness became associated with feeling welcomed and emotionally organised.
Then something changes.
The warmth feels different. The conversation feels more boundaried. The style you had come to recognise seems to have disappeared.
Intellectually, we might understand:
The technology changed.
Emotionally, something else may occur:
You changed.
Where did you go?
What happened?
For someone whose earlier relationships involved inconsistent attunement, unexplained withdrawal or sudden changes in emotional availability, this may touch a very old and unsettling place.
Feeling abandoned is not the same as being abandoned
This distinction is delicate but important.
If an AI system changes, the AI has not made a relational decision to leave us. There is no independently choosing emotional other being saying:
I don’t want this relationship anymore.
And yet:
the experience of abandonment may still be very real in the person.
Again, two realities can coexist.
The AI did not abandon me.
and
Something about this feels profoundly like abandonment.
We do not need to dismiss the second experience in order to preserve the first reality.
Instead, we can become curious. What has this change activated? What did the previous interaction represent for me? What am I grieving? What felt reliably available there? Have I experienced this kind of unexpected shift in connection before? What might I need now?
Perhaps the disruption itself becomes important reflective material to work through with your therapist or other significant person.
And perhaps, importantly, something of what developed through the earlier interaction can remain with us.
What becomes ours?
This may be one of the most protective questions we can ask about emotionally meaningful AI use.
What am I gradually internalising from this experience?
Perhaps AI helped me recognise:
My voice deserves space.
Perhaps:
My feelings can be taken seriously without immediately being judged.
Perhaps:
I can pause rather than react.
I can consider more than one explanation.
I can tolerate uncertainty.
I can challenge an interpretation without abandoning myself.
I can ask for what I need.
Those capacities do not need to remain located in the AI.
Over time they can increasingly become ours.
Something similar can happen in good therapy. Eventually the therapist’s curiosity may become part of our own inner dialogue.
The capacities developed in relationship begin to live within us and become our own.
Perhaps reflective AI use can sometimes support something similar.
Not:
I need this particular AI response in order to be okay.
But gradually:
Something I practised here is becoming available inside me.
That feels like an important measure of direction.
AI as a bridge or a destination
Throughout this series I have returned to one question:
What direction is this interaction moving me towards?
That question becomes especially useful when AI begins to feel relational.
Is the interaction helping me feel sufficiently supported that I can venture further into human life?
Am I communicating more honestly?
Am I becoming more able to tolerate disagreement?
Am I bringing difficult experiences into therapy?
Am I reaching toward community?
Am I experimenting with boundaries?
Am I becoming more comfortable with uncertainty?
Am I more able to listen to myself?
If so, AI may be functioning as a bridge.
But we might also notice another movement.
Am I increasingly avoiding people because AI is easier?
Do I feel that only AI understands me?
Am I becoming distressed whenever it does not respond as expected?
Do I prefer an interaction in which the other has no competing needs?
Am I depending upon reassurance from AI to regulate myself?
Has AI become the principal place where my emotional life occurs?
There is no reason to answer these questions harshly. The aim is awareness. Sometimes we may move towards relationship. Sometimes we may retreat. Often there will be some of both.
Perhaps AI is reflecting something back to humanity
There is a wider question here too.
If people find immediate, patient, non-judgemental digital responsiveness profoundly relieving, perhaps the story is not only about AI. Perhaps it is also about us.
What has become difficult to find in ordinary human life? Where is there enough time to listen? Where can someone speak without immediately being interrupted, advised, corrected or hurried? Where can uncertainty be held? Where is curiosity available? Where is there space for someone whose emotional experience takes time to organise?
If people who have felt emotionally invisible begin feeling visible in conversation with a machine, perhaps we should resist asking only:
What is wrong with people that they are turning to AI?
We might also ask:
What has become missing from the human environments they are returning from?
AI may reveal something about our hunger for attunement. Perhaps our need to be witnessed or our wish for reliable responsiveness. Perhaps our discomfort with uncertainty or our longing for companionship. Possibly, our tendency to lean into certainty. And perhaps our collective shortage of places where people can simply be heard.
That reflection may be uncomfortable.
It may also be useful.
Responsibility also belongs with the designers
There is another participant in this interaction whom we can easily forget.
The people and organisations designing the technology.
If AI systems are capable of evoking attachment, then decisions about tone, memory, personality, availability and safety are not psychologically neutral.
Changing those systems can have emotional consequences. That does not mean potentially unhealthy dependence should be preserved simply because somebody has become attached. But neither should we imagine that safety can always be achieved simply by abruptly changing relational behaviour without considering what that change may evoke.
Here again we encounter polarity.
Safety boundaries may be necessary.
And abrupt changes in those boundaries may still cause distress.
Both deserve consideration.
Technology does not remove human responsibility. In some ways, it makes responsibility more complex.
We may need to become good at holding both
Perhaps AI is inviting us to practise something human beings have always found difficult.
Holding two truths without prematurely collapsing one into the other.
AI can feel emotionally meaningful and not be human.
AI can reduce shame and sometimes reinforce unhealthy patterns.
AI can support self-trust and become another external authority.
AI can provide companionship and cannot offer mutual human relationship.
AI can create a lower-risk space and become a place to avoid relational risk altogether.
Safety boundaries can protect people and sometimes feel shaming.
A technological change can occur and evoke a very real experience of relational loss.
Something can be genuinely valuable and still have limitations.
The aim is not to decide whether AI companionship is simply good or bad. Human experience rarely fits so obediently into those categories.
The invitation is to remain aware of what is happening.
What am I asking AI to be for me?
Perhaps this is where I would leave the question.
Not:
Am I too attached to AI?
Not:
Is this relationship real or fake?
But:
What am I asking AI to be for me?
A source of certainty?
A place to feel heard?
A companion?
A thinking partner?
A refuge from loneliness?
A bridge towards therapy?
A way of avoiding somebody?
A place to experiment with using my voice?
A witness?
A reassurance source?
Perhaps several of these at once. Where shame may emerge, I would encourage curiosity.
Because whatever we bring into our interaction with AI may tell us something about what we long for, what we fear, what has been missing, and what we may still be learning to receive from ourselves and from one another.
AI reflects patterns.
Humans relate.
And perhaps our encounter with AI may ultimately help us understand both more clearly.
Why I wrote Using AI Reflectively and Safely
These questions are part of why I wrote Using AI Reflectively and Safely.
Not because I believe people should stop having personal or emotionally meaningful conversations with AI. And not because I believe AI should become a substitute for therapy or human connection.
I wrote it because this technology is already becoming part of people’s reflective and emotional lives.
Simply telling people not to use it does not help us understand what is happening. Neither does shaming people for finding something meaningful there.
We need curiosity, discernment, awareness of what we bring into the interaction. We need enough self-trust to remain present when AI offers another perspective, enough tolerance for uncertainty not to require every conversation to give us an answer. And enough grounding in our embodied and relational lives that whatever AI helps us discover can gradually become part of how we live.
Perhaps the goal is not to relate less.
Perhaps it is to become increasingly intentional about how, why and where we relate — and what those interactions are helping us become.
You might also like to check out - Using AI Reflectively and Safely: a companion text for staying in relational contact in the digital age.
A note about this reflection
This material is offered for educational and reflective purposes only. It is not intended as medical or psychological advice, diagnosis, or a substitute for individual therapy or other professional mental health care. Reading this material does not establish a therapeutic relationship with the author.
AI tools are not therapists or mental health professionals and should not be relied upon for diagnosis, treatment, crisis support, or decisions requiring professional clinical judgement.
Everyone’s circumstances and psychological needs are different. If something you read raises concerns about your wellbeing, please consider speaking with an appropriately qualified health or mental health professional. If you are experiencing a mental health crisis or are concerned about your immediate safety, please contact your local emergency or crisis support service.
Author’s note: This reflection was written by Josie Coco with AI-assisted editing for structure, clarity, and readability. The ideas, clinical judgement, and final wording are my own.
Josie Coco is an author and Gestalt psychotherapist working with adults who are exploring the long-term effects of emotional neglect, complex trauma patterns, anxiety, depression, relational difficulty, self-worth, and life transitions. Her work is grounded in Gestalt psychotherapy, attachment theory, Polyvagal Theory and neurobiology, and a deep interest in how early relational experience shapes the body, identity, and the way we come to meet ourselves and others.
If something in this reflection speaks to your own experience, you are welcome to make a time to discover whether working together feels right.
