What 6 months of AI journaling taught me about my anxiety patterns

I've been in therapy on and off for most of my adult life. I'm not skeptical of it — I think it works, for me and for many people. But there's a specific limitation I kept running into: the week between sessions is long, and a lot happens in it. By the time I sat down with my therapist, I'd often smoothed over the rougher moments, narrativized the difficult feelings into something more coherent and less raw than they actually were.

Journaling was the obvious supplement. I'd tried it before, always with the same result: three weeks of daily writing, then silence, then a return to the notebook six months later with vague guilt about the gap.

Six months ago, I started using Solen. I didn't expect it to be different. I thought I was just trying another app.

The First Thing I Noticed

The first few weeks felt similar to every other journaling attempt. The AI reflections were good — better than I expected, more specific and less generic than I'd feared. But the thing that surprised me wasn't the quality of individual responses. It was what started happening around week six.

I wrote an entry about feeling behind. Behind on a project, behind on a conversation I'd been meaning to have, behind on some vague undefined standard I couldn't quite name. The reflection came back, and then there was a follow-up question I didn't expect: *You've written about this feeling of being behind a few times now. I'm curious whether the feeling tends to arrive at specific times — is there a pattern to when it shows up?*

I sat with that for a minute. I hadn't told the system anything directly about when I felt this way. But it had read enough entries to notice something I hadn't.

I went back and looked. The feeling of being behind — that low-grade ambient dread — appeared in my entries with unusual consistency on Sunday evenings and Monday mornings. I'd written about it on a Sunday in week two, a Sunday in week four, a Monday in week five. I'd never connected those entries because I'd never read them together.

Sunday Anxiety Has a Name

It turns out what I was experiencing has a documented name: "Sunday Scaries" or anticipatory anxiety, a phenomenon well-enough established that it's generated its own literature in occupational psychology. The dread is anticipatory — it's about the week ahead, not the present moment. It's also, notably, often disproportionate to what's actually scheduled. The anxiety doesn't correlate with how difficult the coming week actually is. It correlates with the end of the weekend regardless.

Knowing this was useful in a different way than simply being told by my therapist that I had anxiety about work. The pattern was visible to me, in my own words, across weeks. I had written myself into a corner and could now see the shape of the corner.

Work Avoidance Was Harder to See

The next pattern took longer to surface. Around month three, I wrote a fairly unremarkable entry about having spent most of the day on low-stakes tasks — clearing email, reorganizing files, doing the kind of work that feels productive but isn't. I noticed I'd written about this before. The reflection noted it too, gently, without making me feel judged: *It sounds like today might have been one of those days where forward motion felt risky somehow. What was the thing you weren't doing while you were doing all the other things?*

That question hit harder than anything a blank page had ever surfaced for me.

When I looked back across three months of entries, I could see a cycle I'd been in for years. Anxiety would spike — sometimes visibly, often not. I'd respond by filling my time with tasks that were real but low-stakes. The difficult project, the difficult conversation, the difficult decision would get backgrounded. Days would pass. The anxiety would temporarily reduce because the threatening thing wasn't being approached. And then it would spike again because I was further behind.

I'd described this cycle to my therapist. She'd named it. We'd worked on it. But seeing it in my own entries — fourteen instances over three months, in my own words — created a different kind of understanding. Not intellectual. Visceral.

What Distance Does

The psychologist Ethan Kross at the University of Michigan has done extensive research on what he calls "self-distancing" — the practice of observing one's own experience from a slight remove rather than being fully immersed in it. His research shows that self-distancing reduces emotional reactivity, supports more constructive thinking about difficult situations, and improves decision-making under stress.

Reading my own entries from two months ago is a form of self-distancing. I can see past-me more clearly than I can see present-me. The feelings are the same, but the charge is lower. And when an AI system surfaces a pattern across those entries — when it says, essentially, "I've watched you write about this thing ten times, and here's what I notice" — it accelerates that distancing effect.

It doesn't replace therapy. It doesn't replace medication for people who need it. It doesn't replace the relational experience of being genuinely witnessed by another person who cares about you.

But it does something else, something harder to get elsewhere: it remembers. And after six months, what I have is not just a collection of entries. It's a record of who I've been, what I've been afraid of, and — slowly, imperfectly — what I've been learning.

That turned out to be worth more than I expected.