Measure against your own past, not other people’s metrics. Kenneth Stanley’s Why Greatness Cannot Be Planned argues real progress appears as new capabilities and new options, so I run a monthly Stepping-Stone Review over my journal — with an AI layer like Mindsera ($14.99/month) or Rosebud ($12.99/month) surfacing the patterns I would miss on my own.
This question shows up in my own search logs, and I recognize the itch behind it. High achievers outsource their sense of progress to external benchmarks — revenue, headcount, followers, titles — and then feel unmeasured the moment they work on anything those numbers cannot see: judgment, focus, recovery, taste. The usual advice is to journal about it, which is where most people stall, because a journal without a measurement method is just a mood archive.
The method I use comes from an AI researcher. Kenneth Stanley’s Why Greatness Cannot Be Planned makes a precise argument: ambitious objectives deceive, because the stepping stones to anything great rarely resemble the destination — and the reliable compass is novelty, which compares your present not to an idealized future but to your own recorded past. His novelty-search algorithm solved a maze 39 times out of 40 by seeking only new behavior; the objective-chasing version managed 3 out of 40. Translated to a person: the honest measure of growth is not distance to a goal. It is what you can do, see, and ask now that you could not thirty days ago.
The Stepping-Stone Review — a monthly protocol
- Reread or export the last month of journal entries and notes, and ask your AI one question: what appears here that did not exist a month earlier — skills, ideas, relationships, questions? Genuine novelty only, no goal language.
- List those as stepping stones collected, and note which ones opened further doors. A stone that spawned three new threads outranks any milestone you merely predicted.
- Rate each thread by pull — what do you actually want to explore more? Stanley’s interestingness is a compass reading, and declining pull is real data about a dead end.
- Finish with one written sentence: “What can I do now that I couldn’t 30 days ago?” If the sentence is empty two months running, the problem is inputs, not measurement.
The journal is the instrument; AI is what makes it readable at scale, because a year of entries defeats human rereading. Two current tools are built for exactly this. Mindsera’s Genius plan at $14.99 a month (or $129 a year) layers 50-plus mental-model frameworks and a proprietary emotion model that connects specific thoughts to specific feelings — the closest thing I have found to pattern detection over your own thinking. Rosebud, at $12.99 a month (or $107.99 a year) for its Bloom plan, takes the conversational route — CBT- and ACT-grounded guided sessions — and has real momentum behind it: a $6 million seed round led by Bessemer and over 150,000 users. And the no-app option works fine: I export raw monthly notes into Claude and run the four questions above as a prompt. The tool matters less than the ritual.
The failure modes deserve naming, because I have hit all three. AI summarization sands off the weird bits — and in Stanley’s terms the weird bits are the stepping stones, so instruct the model explicitly to surface anomalies rather than themes. The apps’ own streaks, scores, and badges are external benchmarks sneaking back in through the side door; a 90-day streak measures compliance, not growth. And every current model flatters by default — ask it to argue that you have not grown this month and see what survives. When the flattery is stripped away and the sentence still fills in, you can trust it.
One more honest limit: some things should stay externally benchmarked. Cash, health markers, whether the product shipped — Stanley’s argument is about ambitious open-ended growth, not accounting. The coaching insight underneath is that benchmark envy is a compass calibrated to someone else’s map. Measured against your own past, with an instrument that cannot lie to you for long, growth without external benchmarks is not unmeasured. It is just measured in the only frame where the number means anything.
Sources: Mindsera’s Rosebud-vs-Mindsera comparison (2026 pricing), rosebud.app, MyLifeNote AI journaling apps compared (2026).
Related: How to Find Your Passion · Best Self-Improvement Books · How to Make Better Decisions · AI Coach App — Building It in 8 Hours