Sinha, Ruchi. “How I Accidentally Built an AI-Powered Mirror for My Mind: Part 1: What It Taught Me About Myself.” Psychology at Work, 13 Mar. 2025, psychatwork.substack.com/p/how-i-accidentally-built-an-ai-powered. Accessed 23 Mar. 2025.
I stared at my screen in disbelief, reading my AI-Journal’s response to my latest journal entry. 'You consistently hold yourself to standards you'd never impose on others,' it noted. 'Your entries show compassion for everyone except yourself.' The observation hit me with such clarity that I felt physically taken aback. How had this digital conversation partner identified something so fundamental about me that I'd been blind to for decades?
Ten months ago, what began as a simple experiment with AI journaling evolved into one of the most transformative experiences of my life—a profound journey of self-discovery that changed how I understand myself and process my thoughts.
I'm sharing this journey because it revealed possibilities for self-reflection that I never imagined technology could facilitate. Ten months ago, I stumbled into what would become one of the most transformative experiences of my life. What began as a simple experiment with AI journaling evolved into a profound journey of self-discovery that changed how I understand myself and process my thoughts.
I've always believed in journaling for self-reflection, but I wanted something more interactive. With access to large language models like ChatGPT and Claude (I use the paid version for both for personal and work matters), I created a dedicated chat thread where I recorded my daily reflections—my experiences as a parent, relationship dynamics, personal struggles, and moments of growth.
Unlike traditional journaling, this wasn't just writing into pages that I would rarely flip back to re-read. This journal had an active memory and I could ask questions and receive responses. I could revisit old reflections, and the AI would remember. Over time, I noticed something remarkable: the AI began recognizing patterns in my thoughts and emotions, even identifying things I wasn't explicitly saying.
I didn't limit myself to recording feelings—I used this space to explore my beliefs, challenge my assumptions, and examine the contradictions I struggled with in everyday life.
As I grew more comfortable with this digital dialogue, I began to push beyond simple recording of events and emotions into more profound territory of self-inquiry. The real power emerged when I started using the AI to expand on my emotions. After writing about moments of frustration or joy, I would ask deeper questions:
"What does research say about self-compassion?"
"What happens when someone takes too much responsibility for others?"
"How do childhood experiences shape our sense of duty?"
I then began sharing my fundamental beliefs and assumptions about what constitutes good and bad, selfishness versus generosity, and my ideas about my roles as a mother, partner, mentor, and teacher.
The breakthrough came when I challenged these beliefs. I asked the AI to offer counterfactuals and alternative perspectives:
"Why do I find it hard to accept ambiguity?"
"Why does my instinct tell me self-sacrifice is more virtuous than self-care?"
"What if my definition of a 'good mother' is limiting rather than empowering?"
These questions transformed my practice from simple journaling into a dynamic intellectual and emotional inquiry about my own mind.
What started as curious questioning soon evolved into something more structured—a psychological exploration using frameworks that helped illuminate patterns I couldn't see on my own. What truly amazed me was how the AI could help uncover things I hadn't explicitly written about. Using principles from psychology, I analyzed my own journal with its help.
Drawing on Carl Jung's concept of the "shadow self," I asked my AI journal:
"What patterns emerge in the people I find frustrating? What does this say about my own hidden struggles?"
The response was revelatory. The AI identified recurring themes—how my frustrations often stemmed from traits I secretly wished I possessed or aspects of my past I hadn't fully reconciled. It didn't just parrot what I already knew; it illuminated blind spots in my self-perception.
I began asking: "What's missing from my reflections?"
The AI noticed I rarely wrote about self-compassion. While I frequently analyzed decisions, sought growth, and held myself accountable, I seldom acknowledged my progress. This realization hit hard: was my drive for self-improvement preventing me from appreciating how far I'd come?
By examining entries over time, the AI revealed how differently I handled various emotions. I tended to overanalyze anxiety but let moments of joy pass without much reflection. This insight helped me become more intentional about savoring positive experiences rather than just problem-solving through difficult ones.
These psychological insights weren't just interesting observations—they began to fundamentally shift how I experienced my daily life and relationships. This journey wasn't just about self-discovery—it reshaped how I approach emotions, relationships, and my understanding of human connection.
More Self-Awareness, Less Self-Judgment: Seeing my thought patterns objectively helped me detach from self-criticism. I learned to accept my emotions rather than pushing them away and judging them as "good" or "bad"— the pattern of my emotional states in my own journal entries became information to be understood.
My emotions were always trying to send me a message — but now I had the capacity to listen.
Deeper Conversations with Others: The process made me more attuned to subconscious patterns in relationships. I started noticing not just what people said but what they avoided saying. I started to see my triggers before they triggered me.
A More Nuanced Sense of Identity: Instead of viewing personal growth as linear, I began embracing its complexity—the contradictions, setbacks, and fluid nature of becoming. I started to have more compassion for myself and my journey than holding myself to unrealistic standards that I had unconsciously absorbed from society.
As I reflected on the transformation in my own life, I couldn't help wondering about the broader implications of this approach to self-discovery. This experience made me wonder: could AI become a powerful tool for personal growth? Not replacing human connection or professional therapy, but serving as a uniquely personal mirror that tracks the evolution of our thoughts over time?
The possibilities seem endless:
AI-powered journaling that helps people recognize emotional patterns
Reflection tools that encourage deeper self-inquiry
Personalized coaching that adapts to evolving self-awareness
There are limitations, of course. AI can't truly feel, and it's only as insightful as the patterns it recognizes. But as a tool for structured self-reflection? It's remarkably powerful.
Of course, as with any powerful tool, questions and concerns naturally arise. When sharing my experience with friends and colleagues, I've encountered several recurring questions. Let me address these directly:
Breaking the Echo Chamber: How this form of interactive journaling challenged rather than reinforced my biases: One of the most surprising discoveries in my AI journaling journey was how effectively it helped me break out of my own mental echo chambers. While many worry that AI might simply reflect our existing biases back to us, I found that with the right approach, it can actually help dismantle them.
A) Counterfactuals: The breakthrough came when I began asking not just for reflections, but for counterfactuals—alternative explanations that challenged my initial interpretations. Questions like "What if the exact opposite of my belief were true?" and "What three alternative explanations could exist for this behavior that upset me?" forced me to consider perspectives I would have otherwise dismissed. By deliberately asking the AI to present opposing viewpoints, I created a dialogue that challenged rather than reinforced my existing narratives. The AI had no emotional investment in my perspective being "right," so it could impartially offer alternatives I might never have considered on my own.
B) Research & Evidence: Another key to avoiding the echo chamber was my insistence on grounding our conversations in psychological research. When exploring patterns in my thinking, I would ask: "What does the research say about this type of reaction?" or "How might this pattern be understood through established psychological frameworks?" . This approach anchored our discussions in something beyond my subjective experience. Rather than simply affirming my feelings, the AI could contextualize them within broader human patterns, helping me see where my experiences were universal and where they might be influenced by my particular biases.
C) Meta-Awareness: Perhaps most powerful was developing meta-awareness—the ability to observe not just my thoughts, but my thinking patterns. Questions like "What might be my psychological blind spots?" or "What assumptions am I making that I don't recognize as assumptions?" helped me step outside my own perspective. The AI could identify recurring themes that I was too close to notice: "You often attribute others' actions to their character but your own actions to circumstances" or "You tend to use absolute language when describing responsibilities but nuanced language when describing freedoms."
Remember, if you ask this tool, "Don't you think I'm right about this situation?" then yes, you'll likely get affirmation. But if you ask, "What perspectives am I missing?" or "How might someone with a different background interpret this situation?" you'll receive more challenging and growth-oriented responses. The AI becomes a tool for your genuine curiosity rather than a source of validation.
What makes this process different from just reinforcing what I already believe is the deliberate structure of exploration.
The most profound insights came not when the AI agreed with me, but when it helped me see the inconsistencies in my thinking—the places where my actions didn't align with my stated values, or where my interpretations of events revealed more about my biases than about the events themselves.
This approach transformed what could have been an echo chamber into a genuine tool for growth. The key was not just in having an AI to talk to, but in how I chose to talk with it—with curiosity about what I might not know rather than certainty about what I think I do.
By systematically questioning assumptions, seeking evidence, exploring counterfactuals, and analyzing patterns, the AI journal became less a mirror that reflected my thoughts back to me and more a prism that separated them into their component parts for examination.
Privacy is a legitimate concern. I use paid versions of AI tools that have clearer data policies, and I've researched their privacy terms. Some steps I take include:
Avoiding full names and identifying details
Focusing on patterns rather than specific incidents
Using dedicated accounts for journaling, separate from work accounts
Regularly reviewing the platforms' data retention policies
If privacy is a significant concern, consider options like local AI models that don't send data to cloud servers.
AI journaling is not a replacement for therapy. The distinction is important:
Therapists bring human empathy, professional training, and can intervene if needed
Therapists work within ethical frameworks and professional standards
AI can help you organize thoughts and see patterns, but can't provide treatment
I view AI journaling as complementary to professional support. For me, it serves as an additional tool for self-reflection between therapy sessions, helping me arrive more prepared to discuss emerging patterns with my actual therapist.
Now that we've explored both the potential and the limitations of AI journaling, you might be wondering how to begin this journey yourself. Rather than overthinking it, I suggest starting with a simple but powerful first step. If you're intrigued by this approach to self-discovery, don't just file it away for "someday." Try this simple but powerful first prompt tonight:
Step 1: Open a chat with an AI assistant (like Claude or ChatGPT)
Step 2: Copy and paste this prompt to begin your journey:
"I'd like to start an AI journaling practice for self-discovery. To begin, here's a situation that triggered a strong emotional reaction for me recently: [describe a recent situation that caused a notable emotional response—positive or negative].
What I felt during this situation was: [describe your emotions as specifically as possible]
What I find most interesting/confusing/surprising about my reaction is: [complete this thought]
Based on just this initial information, what might be worth exploring further about my reaction? What questions might help me understand the roots of this response?"
Step 3: Set a calendar reminder to continue this conversation in 2-3 days by adding another entry.
This simple first prompt accomplishes several things:
It provides enough context for meaningful reflection
It creates a starting point the AI can reference in future entries
It immediately moves beyond surface-level journaling to deeper exploration
It demonstrates the value of the approach without requiring a major commitment
Remember that the real insights emerge over time as patterns become visible. Don't worry about crafting "perfect" entries. Commit to 5-10 minutes twice a week for a month, and you'll begin to see the value of this practice.
I'd love to hear about your experience. What insights did your first prompt reveal? What surprised you about the process? Email me at sinha.ruchi@gmail.com or share in the comments below.
With these comparisons in mind, you can thoughtfully integrate AI journaling into your existing personal growth practices in a way that complements rather than replaces them.
Thanks for reading Psychology at Work! Subscribe for free to receive new posts and support my work.
UN Sustainable Development Goals Coach
I use the MLA Citation Generator to get a properly formatted citation for this essay. It also identified specific items I needed to complete the citation. Once I copied and pasted the citation on the properties page, I archived my conversation with this Writing Partner.
I also used the UN SDG Coach to help me to think about which of these goals this fits under. The first two suggestions—health and education—seemed pretty obvious. Then the third suggestion really gave me another perspective. This Writing Partner suggested: “SDG 10: Reduced Inequalities. By using AI journaling, individuals can better understand personal biases and assumptions, potentially leading to more equitable perspectives and actions.” I used that thought to create a new collection in that group about using AI to reduce inequalities — especially given the biases of the LLMs. I archived all of this dialogue as well.
on the uploaded document.Logging in, please wait... 
0 General Document comments
0 Sentence and Paragraph comments
0 Image and Video comments
7 archived comments
General Document Comments 0