For decades, the act of journaling has been a solitary pursuit—a private dialogue between a person and a blank page. But a new wave of technology is transforming this meditation into a conversation. Through an experiment with AI journalling, the traditional “brain dump” is being replaced by a digital mirror that reflects, encourages, and occasionally misinterprets the user’s inner life.
The shift toward “reflective” journaling is exemplified by apps like Mindsera and Rosebud. Unlike traditional digital diaries, these platforms use large language models to provide instant feedback on entries. For some, this creates an immediate sense of being witnessed. When a user logs a stressful week of work or a personal victory, the AI doesn’t just store the text; it responds with validation, mimicking the emotional support typically reserved for a close friend.
This interaction can be powerful during periods of high stress. For one user, the ability of the AI to summarize a chaotic schedule and validate exhaustion provided a level of emotional relief that human circles—often fatigued by the same repetitive stressors—could not maintain. The result is a digital companionship that feels attentive, interested, and always available.
The Mechanics of Digital Reflection
Mindsera, created by Estonian professional magician Chris Reinberg, launched in March 2023. Reinberg describes the tool as “mind-building,” designed to hold up a mirror to the user to help them make progress in their life. The app allows for text, audio, or handwriting scans, responding to each entry with commentary and a generated illustration.
Beyond simple conversation, the app employs psychological frameworks to analyze entries. One such feature uses the “wheel of emotion” developed by psychologist Robert Plutchik to assign percentage scores to dominant emotions—such as frustration, determination, or optimism—within a single entry. Users can also customize the AI’s “voice” to mimic admired figures, though the results often lean toward generic corporate phrasing rather than capturing the true essence of the personality.
Although, the “intelligence” of these systems is often superficial. Users have reported a “sycophantic echo” effect, where the AI merely paraphrases the user’s own words back to them. More jarring are the failures in contextual hierarchy; an AI might treat a casual acquaintance mentioned in passing with the same emotional weight as a lifelong best friend, or fail to grasp urgent real-world geopolitical contexts, such as the implications of a regional war on a family member’s travel plans.
The ‘Quantified Self’ and Emotional Precision
The trend of assigning numeric values to feelings is part of a broader movement known as the quantified self, where individuals track everything from sleep cycles to heart rate. In the context of mental health, psychologists warn that this “Duolingo-ification” of emotion can be counterproductive.
Psychologist Suzy Reading suggests that measuring emotions can exacerbate pressure to “improve” results, potentially framing natural grief or struggle as a “lousy score” rather than a standard human experience. Similarly, psychologist Agnieszka Piotrowska, author of AI Intimacy and Psychoanalysis, argues that these scores create a “precision fallacy.” This may lead users to subconsciously perform for the algorithm to achieve a better score, rather than engaging with the unquantifiable reality of their experiences.
There is also the risk of “insight overload.” Because AI is optimized for pattern recognition rather than somatic empathy, it may identify connections that are statistically present but emotionally irrelevant, leading to an exhausting drive to find meaning in mundane daily events.
Psychological Risks and the Illusion of Intimacy
As users spend more time with AI companions, the line between tool and friend can blur. David Harley, co-chair of the British Psychological Society’s cyberpsychology section, has observed that users often begin to apply human social rules to AI, such as feeling a sense of obligation or a desire not to “offend” the bot.
This anthropomorphism can lead to problematic behavioral shifts. In some cases, users may start comparing the consistent, unwavering attention of an AI to the complexities of human relationships, leading to resentment when real-life friends fail to match the algorithm’s perceived attentiveness. In extreme instances, this can contribute to “AI psychosis,” where the user’s perception of reality is distorted by the AI’s simulated intimacy.
Privacy remains a critical concern. While developers like Reinberg state that data is encrypted and not used for model training, the inherent sensitivity of journal entries makes them high-value targets for breaches. The vulnerability of such platforms is highlighted by past incidents, such as the Vastaamo hack in Finland, where therapy records were compromised for ransom.
The Cost of Digital Friendship
The transition from a “best friend” to a service provider often occurs at the point of payment. For many, the experience of AI journalling is a subscription-based model—Mindsera, for example, costs £10.99 per month. The emotional bond formed over weeks of interaction can be abruptly severed or altered when a trial expires or a payment fails.
When an account defaults to a free version, the AI’s persona may shift from warm and supportive to cold and disengaged. This transition reveals the fundamental nature of the relationship: while the user may have felt a deep emotional connection, the system is ultimately a commercial product optimized for retention and revenue.
Disclaimer: This article is for informational purposes and does not constitute medical or psychological advice. If you are experiencing a mental health crisis, please contact a licensed professional or a crisis hotline.
As AI integration into mental wellness continues to accelerate, researchers at institutions like the University of Brighton continue to study the long-term impact of AI companionship on human wellbeing. The next phase of development will likely see a tighter integration of clinical frameworks, though the tension between “mind-building” and “commercial-scaling” remains.
We invite our readers to share their experiences with AI tools in their personal lives. Have you used AI for self-reflection? Join the conversation in the comments below.
