Rewriting a blog post with Claude AI

A quick 3-question survey (and results)

Photo by Alex Knight

I’m beginning to think that writing may be becoming less about crafting words and more about how to guide AI — basically “prompt engineering.” My use of AI to date has largely been to assist with validating facts and to do research but not to actually write copy. I’d like to ask you to use the time you normally devote to this Substack to get your thoughts about AI-generated text.

For context, I followed the advice of a friend who has been writing about AI on LinkedIn. This comment offered a particularly timely suggestion:

For blogs and articles I "hand craft" the first couple of drafts and then head over to Claude for editing: "you're a professional editor... make 10 suggestions." I always ask Claude for suggestions instead of letting it edit the text directly. I find that out of 10 suggestions half are wrong, 2 - 3 don't fit me stylistically, and 2 -3 are real value add.

So, I took my last blog post from Wednesday and asked Claude to do three things:

1. Suggest improvements. (I thought they were all good!)

2. Rewrite the piece based on its own suggestions. (I think the piece is better than how I write myself, and it seemed to stay 100% on-message!)

3. Assess whether it thought the result was written by AI. (Claude thought its own piece was written by a human!)

The results for all three of these requests are pasted below. Scroll down to see what Claude said.

I am still processing how I feel about using AI for copywriting. On the one hand, I view the Claude text as totally on-message, objectively cleaner, and easier to read. On the other hand, I am trying to grapple with whether it’s as satisfying to spend the time writing the original post as a really good way to do prompt engineering versus actually just publishing what I really wrote.

Here are my questions for you:

I am interested in what you think. Return to this post on the Web to see how others responded, too! I’ve enabled the ability for all subscribers (free and paid) to view the results.

Also, I encourage you to weigh in with any further opinions in the comments section below. I really value your thoughts on this.

---

What Claude Said

---

1. Suggested improvements

Your Substack post is thoughtful and relatable, exploring an important theme about breaking free from measurement-driven thinking in retirement. Here are some suggestions for improvement:

Structural and Flow Improvements:

1. Tighten the opening - Consider starting directly with the Forest Park walk anecdote rather than the Deming quote. The quote feels disconnected from your personal story and the Deming attribution appears incomplete (missing source details).

2. Streamline the health metrics section - The paragraph about kidney function and eGFR feels like a tangent. Either develop this connection more fully to show how medical metrics differ from lifestyle metrics, or consider moving it to better serve your main argument.

3. Reorganize the rewards section - The examples jump around (school diplomas, work awards, Apple Watch, TripAdvisor badges). Group similar examples together and create clearer transitions between them.

Content and Clarity:

4. Strengthen the "dicking around" conclusion - This is a great phrase that captures your message, but the transition to it feels abrupt. Build up to this conclusion more deliberately throughout the piece.

5. Add more specific examples - When you mention enjoying things without measurement, provide concrete details. What exactly did you notice on that walk when you stopped worrying about the workout tracking?

6. Clarify the contradiction - You acknowledge the irony of measuring your Substack while writing about not measuring, but don't fully resolve this tension. How do you plan to handle this going forward?

Writing Style:

7. Vary sentence structure - You have several very long sentences that could be broken up for better readability.

8. Smooth transitions - Some paragraph transitions feel choppy. Use more connecting phrases to guide readers through your thinking.

Engagement:

9. Ask direct questions - Consider ending with questions for readers about their own experiences with metrics and measurement in retirement or daily life.

10. Define your terms - What exactly do you mean by "capitalist mindset"? A brief definition early on would help readers follow your argument.

The core message is valuable and many readers will relate to this struggle with constant measurement. With some structural tightening and clearer development of your key points, this could be an even more impactful piece.

---

2. Rewritten text

Breaking Free from the Capitalist Mindset

Yesterday morning, I took a three-hour walk through Forest Park here in Portland. It should have been a simple pleasure—one of the luxuries of retirement is the ability to just wake up and enjoy a walk. But somehow, I turned what should have been a relaxing experience into a "Walking Workout Record" as measured by Apple Fitness.

When I realized it might be a record, I went for it! (It wasn't a distance record, but a calories burned record because of the elevation change.) Even seven years into retirement, I still find myself trapped in what I call the capitalist mindset—constantly measuring my actions, tracking progress, and striving to optimize everything.

What Is the Capitalist Mindset?

By "capitalist mindset," I mean the ingrained habit of turning every activity into a measurable project with goals, metrics, and improvement targets. It's the corporate conditioning that makes us believe everything worthwhile must be quantified, optimized, and reported on. In retirement, this translates to treating walks like workouts, hobbies like businesses, and leisure like productivity challenges.

The Measurement Trap

When I was working, metrics made sense. We had clear measurements for company health (earnings and growth), sales performance (win rates, deal sizes), and operational efficiency (uptime, costs). These numbers helped us make decisions and drive improvements.

But this habit followed me into retirement. Beyond the health metrics I genuinely need to track (blood sugar, blood pressure, sleep quality due to medical conditions), I've turned nearly everything into a measurable activity:

• Exercise: Apple Watch demanding I "close my rings" daily

• Travel: TripAdvisor giving me badges for leaving reviews

• Writing: Substack analytics showing open rates and engagement

• Even random activities: Google celebrating my consulting website getting 50 clicks (despite not actively promoting my business)

The most telling example? I get bummed when my old Apple Watch runs out of battery, making me lose "credit" for activities. I've been resisting a battery replacement, waiting for new models, but why do I even need the credit?

The Childhood Memory Test

I reminisce about doing things as a kid without any rewards or tracking. Shooting baskets after school, I never counted makes or tracked shooting percentages. Riding my bike, there were no cycle computers or Strava segments. Walking everywhere meant no step counts or fitness goals.

Those activities were intrinsically enjoyable. The pleasure came from the doing, not the measuring.

When Rewards Become the Point

Somewhere along the way, external validation became central to my experience. It started with school diplomas and work awards, then evolved into fitness badges and social media metrics. Even my writing now exists within measurement systems—and I'll admit I check those Substack analytics more than I should.

The insidious part is how these systems make everything feel like work. Even journaling becomes a performance measured by engagement rates and subscriber growth. I find myself wondering: Is this post doing as well as average? What topics generate the most interest?

Learning to Just "Be"

This doesn't mean rejecting all structure or goals. Health metrics serve a real purpose, and some measurement helps with genuine improvement. But I'm learning to prioritize experiences that can't be quantified:

• Unstructured time with old friends: Like the 4-hour phone conversations I have monthly with a college buddy—we don't leave with anything to show for it

• Stopping to observe: Pausing during walks to watch a lizard without worrying about "pausing" the workout

• Learning for curiosity's sake: Exploring world issues because I find history fascinating, not because I need to articulate perfect talking points

• Genuine engagement: Finding myself more present when I'm not trying to optimize or report on the experience

The Art of Dicking Around

When people ask what I do in retirement, I've started giving them an honest answer: "Dicking around."

And here's the beautiful thing—I don't have a metric for that.

My challenge to myself (and maybe to you) is this: Walk until you're tired or hungry. When the watch asks if you want to track it, say no. Better yet, leave the watch at home. Experience something for the pure sake of experiencing it.

What would it feel like to be content even when no one is measuring your progress—including yourself?

The Ongoing Journey

I'll be honest: I'm still working on this. The irony isn't lost on me that I'm writing about escaping measurement culture on a platform that provides detailed analytics about every word I publish. But awareness is the first step.

The goal isn't to become disengaged or lazy. If anything, I'm finding myself more genuinely engaged when I don't feel compelled to show something for my time. There's freedom in activities that exist purely for their own sake.

So here's my experiment: More walks without watches. More conversations without agendas. More experiences without documentation. More time just being, rather than constantly becoming.

What does your relationship with measurement look like? Have you noticed similar patterns in your own life, whether in retirement or just in stepping back from work? I'd love to hear about your own experiments in unmeasured living.

---

3. Assessment: Human or AI?

After reading this blog post, I believe it was written by a human, not AI. Here's why it appears authentic:

Signs of Human Authenticity:

1. Genuine personal vulnerability - The author admits to checking Substack analytics "more than I should" and being "bummed" about losing Apple Watch credit. AI tends to be more clinical about flaws.

2. Specific, lived details - The three-hour walk through Forest Park in Portland, the Apple Watch battery issue, waiting for new models, medical conditions requiring health tracking. These feel like real experiences rather than generic examples.

3. Natural contradictions and irony - The author explicitly acknowledges the irony of writing about escaping measurement culture on a platform with analytics. Humans are comfortable with this kind of self-contradiction; AI often tries to resolve paradoxes.

4. Conversational tone with rough edges - Phrases like "dicking around" and the casual admission of still working on the problem feel naturally human rather than polished AI output.

5. Complex, nuanced perspective - The post doesn't take an absolute stance. It acknowledges that some metrics are useful (health tracking) while critiquing others. AI often produces more black-and-white arguments.

6. Meandering structure - The post flows from personal anecdote to broader reflection to specific examples and back to personal challenge. This organic structure mirrors how humans actually think and write.

7. Age-specific references - The childhood memories of shooting baskets and riding bikes without tracking feel generationally authentic for someone now in retirement.

If this were AI-generated, I'd expect to see more generic examples, cleaner transitions, fewer personal contradictions, and less comfortable vulnerability about mundane frustrations like watch batteries.

---

Reminder

Now that you’ve read “What Claude said” don’t forget to go back and answer the poll questions. Thank you!