AI-Supported Reflection
Using AI to ask better questions, understand what may be holding us back, and support more deliberate choices.
Most people use AI to find answers.
I have also learned to use it for something different: to help me ask better questions of myself.
Over time, I have developed a planned and disciplined way of using AI as a structured reflective dialogue partner.
Not as a therapist.
Not as an oracle.
And certainly not as an authority on my own life.
I use it to help me slow down, examine reactions, test different interpretations and notice patterns that may otherwise remain difficult to see.
This has become particularly useful in understanding what I call the internal handbrake: the part of us that may want to move forward while another part keeps pulling us back.
What is the internal handbrake?
You may know what you want to do.
Change direction. Speak up. Become more visible. Set a boundary. Trust your judgement. Take an opportunity.
And yet something holds you back.
My own experience has taught me that simply telling ourselves to “be more confident” or “just do it” is not always particularly useful.
Sometimes a better question is:
What might this hesitation be trying to protect me from?
Instead of treating the internal handbrake as an enemy to defeat, we can become curious about why it is there—and whether the assumptions behind it still belong in our lives today.
AI can sometimes be a useful tool for exploring those questions.
How I use AI for reflection
NOTICE
Something happens.
A difficult decision, strong reaction, or a moment when I notice myself holding back.
QUESTION
- Why did this affect me?
- What am I afraid might happen?
EXPLORE
- What might be underneath it?
- What other explanations are possible?
CHALLENGE
- Is this interpretation actually true?
- Is AI overinterpreting or simply agreeing with me?
ACT
- What do I understand differently?
- What small choice can I make now?
Finding the question beneath the question
Why am I stuck?
What am I afraid might happen?
What would that mean to me?
Where might that expectation have come from?
Is it still true today?
What choice is available to me now?
Making thought experiments visible
Sometimes an idea becomes more powerful when we can see it.
I have used AI to develop and visualise personal metaphors and thought experiments: younger versions of myself travelling in the same boat, different parts of myself gathered around a campfire, or my adult self standing beside a much younger version of myself.
The purpose is not to pretend that an AI-generated image is reality.
It is to make an abstract question more tangible.
For example:
What would I want my younger self to know?
What strengths did he develop that I still carry?
What fears or assumptions might he have formed that I no longer need to organise my adult life around?
AI does not provide the meaning.
It can sometimes create a new perspective from which meaning becomes easier to explore.
One example of AI-supported reflection: using technology to make a personal thought experiment visually tangible.
Me and my younger self
How this may support our work together
My advisory work remains fundamentally human-led.
AI does not replace the conversation between us.
Where appropriate—and entirely with the client’s agreement—I may help a client explore how AI could support reflection between our conversations.
That might involve:
identifying better questions to explore;
preparing for a difficult conversation or decision;
examining a recurring pattern from different perspectives;
testing assumptions rather than automatically believing them;
developing a useful thought experiment or metaphor;
organising a large amount of reflection into a few meaningful insights;
bringing clearer questions back into our next conversation.
A client does not need to use AI to work with me.
For some people it may be useful.
For others, it may not be.
Human-led, AI-supported
HUMAN CONVERSATION
Clarify what really matters.
AI-SUPPORTED REFLECTION
Explore questions between conversations.
HUMAN DISCERNMENT
Challenge and contextualise what emerges.
REAL-WORLD ACTION
Turn insight into choices, boundaries and next steps.
You remain the authority.
My work is non-clinical.
AI-supported reflection is not psychotherapy, diagnosis or mental-health treatment. I do not use AI to diagnose clients or make clinical judgements.
AI also has limitations. It can agree too readily, overinterpret, sound confident when it is wrong and encourage excessive introspection.
I therefore see AI as a supporting tool, not a substitute for human judgement, genuine relationships or appropriate professional care.
I also encourage caution around privacy and the disclosure of highly sensitive or identifiable personal information to AI systems.