I have kept a private journal for 25 years. 143,000 words in one Word document. I always meant to reread it. I never made the time.
I asked Claude to read the whole thing. In three minutes it handed back a portrait of my fears, my repeated mistakes, my growth and the people who matter most to me. The summary was accurate. Then it did something I did not expect. It recommended how I should spend the next 10 years.
This is the kind of work that sits in the important-but-not-urgent corner and never gets done. AI finished it in one sitting.
I was so impressed by the its recommendations that I saved them as a note on my phone and built a Claude Cowork automation that emails the list to me on the first of every month. The analysis cost three minutes, but the value will compound for years.
The Data
The first entry is dated June 8, 2001 and the the last is April 11, 2026. The journal runs 143,000 words across 630 entries, an average of 24 entries a year. 2002 holds 48, driven by a divorce, a failed startup and a long stretch in a hotel. 2009 and 2012 sit nearly empty. 2016 marks a sustained return to the practice.
The Prompt
As usual, I “asked AI” to write the prompt first. I opened a chat, described the goal and let it produce the instructions. Here is the prompt Claude created:
“I am sharing 25 years of my personal journal. Read all of it before responding. Answer each question using direct evidence from the journal. Cite approximate time periods. Be honest, direct, and specific. Do not generalize. Name any contradictions or surprises explicitly.
Patterns and themes:
1. What concerns or goals appear most often across 25 years?
2. What did I worry about that never materialized?
3. Where do I repeat the same mistakes?
Growth and change:
4. How have my core values shifted?
5. Which early beliefs have I abandoned, and why?
6. Where have I grown most, and what drove it?
Relationships:
7. Which relationships have I invested in, and which have I neglected?
8. How do I respond to conflict, and has that changed?
9. Who appears most often, and what does that reveal?
Decision-making:
10. What decision frameworks do I rely on?
11. Which decisions do I regret most, and what connects them?
12. Where do I act on instinct versus deliberation?
Professional and purpose:
13. How has my definition of success evolved?
14. When was I most energized, and what created those conditions?
15. Where do my stated goals and my behavior diverge?
Blind spots:
16. What do I avoid writing about?
17. Where do I blame outside factors instead of taking ownership?
18. What feedback recurs that I have not acted on?
Close with a one-page summary of the three most important insights about who I am and what I should change.”
I started a fresh chat, pasted the prompt and uploaded the document.
What Claude Found
The model wrote back in the second person. The edited excerpts below are the parts I am comfortable making public. Some findings stay private, which is the point of a journal.
Patterns.
Career and professional identity dominate every period. Unemployed in 2001, consulting in 2018, building GAI Insights in 2026, work is the lens you use to judge your life. Exercise and health appear as a constant aspiration and a rare reality. Fatherhood, and your relationship with your daughter, runs through all 25 years with unusual depth.
Growth.
In 2001 you measured self-worth by financial success. By 2018 you wanted a double-wide with Claudia, a good dog, laughter and high integrity. The move from achievement-as-worth to contribution-as-worth is real, though not complete.
The largest change is emotional regulation. The 2001 Paul was in clinical depression, scored 50 on a depression inventory and was unable to leave a hotel room for days. By 2017, Stoic philosophy reads as genuinely internalized.
Decision-making.
You rely on three frameworks. Stoic acceptance, focused on what you control. A pros-and-cons structure, used well in your October 2018 job analysis. And input aggregation, where you poll friends and mentors, then triangulate. The third one works when you act on the input and fails when you collect it and stall.
Instinct serves you on technology. You moved into the internet, energy management, machine learning and AI advisory before each became mainstream.
Goals versus behavior.
You commit to exercise in almost every entry. The record shows bursts of three to five sessions a week broken by long gaps. You write about rising at 5 a.m. and working outside the house. The record shows aspiration more than habit.
Why This Worked
Good output starts with good data. Twenty-five years of honest journaling is an uncommon data set about one person. Give AI that input, and it returns a mirror most of us never take the time to hold up. The tool matters less than the input. Models from Anthropic, OpenAI and Google all handle a document this size. In this amazing Age of AI, the constraint is no longer the analysis. The constraint is whether you have the data, imagination and impetus to act.
