Hello and welcome to your weekly dose of actionable (and occasionally provocative) things.

My new hobby: trying to justify my Mac mini purchase.

God bless my wife for such timely approval!

Because the config I bought (small SSD, extra memory) is already removed entirely from apple.com.

So this was not an impulse purchase.

This was supply-chain awareness.

Possibly panic.

By the way, if you want a life hack for buying Macs directly from official apple.com for $100 less, reach out!

Anyway, as a few of you probably remember, I run local AI agents.

Or, more accurately, I try to run local AI agents. Important difference.

Until recently, it was OpenClaw.

Then I switched to Hermes.

So far, I like Hermes more because it builds memory and actually uses it better.

Last week, I used it to replicate two paid apps I found useful.

First one: inventory management, properly this time.

The idea:

  1. I pull items out of a box in the garage

  2. I take a picture of each item

  3. I tell the agent: “Add these to Box 1 in the garage”

  4. The agent recognizes what’s in the image

  5. It adds everything to Notion with the image, location, quantity, search keywords, and date

Notion, obviously, because every personal problem must eventually become a database.

This is not because I love inventory. Nobody loves inventory.

I just want to drop something into any random box without thinking about where it belongs, but still be able to find it in no time.

The best part is that I did not need to design the whole thing myself.

I just asked questions.

My first prompt was basically:

“What would you do if I wanted to create an inventory management system where I send you a picture and location, you recognize the item, and then add it to Notion with the image, location, quantity, search keywords, and date?”

Then I asked a few follow-ups:

  • Do we need to split this between different sub-agents?

  • Can we use the existing ones?

  • Is this the best way to optimize token usage?

  • How would you make it more reliable?

  • etc.

And finally:

“Ok, build the flow and analyze it for consistency and optimization.”

Magic!

The second one was even simpler.

I asked:

“How would you build a flow if I wanted to track my calorie intake by photo, inspired by CalAI, but store the results in another Notion database with the nutrition breakdown?”

Because the agent had already built the inventory flow and updated his (it’s a boy!) memory, the result was shockingly good right away.

Pro tip: I added calibration.

At the beginning, I sent a photo of my hand next to a ruler so the AI could learn what size my hand is. Now, when I take a picture of a meal, I include my hand so it can estimate portion size more accurately.

Very normal.

Nothing strange about a grown man photographing two slices of pizza next to his hand for science.

Enjoy the edition!

Table of Contents

Heads-up: If you’re reading this in Gmail (either in the browser or the app), you might not see the full content. Since our emails are packed with rich content, Gmail loves to clip them. You’ll see a small note at the bottom that says [Message clipped] View entire message. Just click that link - the rest of the content is waiting for you on our website.

Learn From My Mistakes

Short story of how I break life chaos into small, solvable problems - 2 min read.

I’ve noticed one trend.

My manager assigns me a task or project → I put in the work.

I put in the work.

Thinking. Organizing. Optimizing.

Making it clear enough that another person can open it and not immediately suffer.

After a week of sweat, I present it to him.

His usual response: “Good starting point… But…”

And I’m like: “O_o. Dude. This is a finished product. Ready for consumption. Enjoy your meal.”

But then he starts asking questions.

Poking the whole thing. Trying to find holes.

Not going to lie, this is not my favorite part of the experience.

My first reaction is usually very mature and professional:

How dare you question my beautiful creature?

But then something annoying happens.

He starts asking questions that actually make sense.

Even worse, they put me in his shoes.

Very presumptuous of me to claim, of course. But still.

I start seeing other angles.

Other risks. Other possible versions of the solution.

And then it takes another bunch of effort.

Next 1:1, same situation.

He says: “Good progress. What if…”

And I’m back to: “O_o.”

The challenge is simple.

We love our ideas because we created them.

We protect them from the hostile environment like tiny emotional houseplants.

But other people do not care.

They have no attachment to the thing we spent so much effort building.

To them, it is just one possible option.

Not one version, but version one.

Fair enough.

I mean, the world is not fair by any means. So maybe “fair enough” is too generous.

But still.

How do we survive this without losing sanity?

The answer is annoyingly simple.

Reframing.

Old attitude: “I love the solution I came up with.”

New approach: “I love the problem I am dealing with.”

That small shift changes everything.

Because a solution is a one-time thing.

Maybe it works. Maybe it doesn’t. (Most likely, it doesn’t. That is not pessimism.
That is just probability being rude).

The value is not in this solution. The value is in being able to make the next one.

Now it is not checkers.

More like 3D chess.

From The Big Bang Theory. Remember?

No? Maybe you are too young to read this newsletter.

Anyway.

That is an entertaining way of saying: there are more angles than I want there to be.

And the rule of this game is boringly simple:

To create good ideas, you need to come up with a lot of bad ones.

And then you need to let the bad ones go.

There is an old parable about the “monkey trap.”

A narrow hole is made in a coconut. A banana is placed inside. The coconut is tied to a tree.

The monkey reaches in, grabs the prize, and can no longer pull its fist out.

In theory, it could free itself at any moment.

All it has to do is let the banana go.

But that is exactly what it does not do.

For those who are not subscribed yet, let me explain:

The banana is the allegory of bad ideas.

Sticking to bad ideas is also a bad idea.

Very advanced stuff here.

Write this down instead:

Fall in love with the problem, not the solution.

Or, to make it even more practical:

Let the banana go!

Till next time.

The Curious Procrastinator relies on word of mouth!

If you’re enjoying our newsletter, please help us reach more readers by forwarding this letter to a friend.

Our favorite digital finds

Tools, apps, and services that actually deliver

This is music discovery as a map and a time machine. It is a good little rabbit hole for songs you probably would never think to search for yourself.

This is a camera app for people who want tourist photos with fewer tourists in them. It shoots a burst of photos, looks for the moving crowd, and tries to give you the landmark without the humans.

If “one quick check” on your phone keeps turning into half an hour, try Momotaro. It runs a timer and blocks the apps you already know you should not open.

Short & Sweet

Short articles worth your attention

The Last Quiet Thing - 8 min read.

An essay about how modern devices create constant work: updates, charging, logins, subscriptions, alerts, and settings. It argues that the problem may not be screen addiction. It may be maintenance.

Are emojis at work harmless warmth, or tiny career-limiting hieroglyphs? This study gives a more useful answer than “it depends,” although, it does depend. If you write in Slack, Teams, or any workplace chat where tone goes to die, this is a practical read.

Humans are funny. If sad songs, tragic novels, or quietly ruinous movies somehow make you feel better instead of worse, this piece explains why.

Add this to your shelf

If you're looking for something to read, this book's worth considering

The book is about using better mental frames to handle stress, motivation, happiness, and success. Scott Adams (yup, the one who created Dilbert) keeps it practical and starts from a useful assumption: people are not nearly as rational as they think.

Feeling the vibe? Drop your email and we will deliver more weekly.

A Workspace I Envy

A handpicked desk setup that caught my eye this week

Everyone gets nostalgic for the time when the sun was brighter, the grass was greener, and the trees were taller. My version is less romantic: floppy disks, tiny monochrome displays, and thinking 1.44 MB was plenty. Yeah, and the walls made of real bricks.

Behind the Persona

A deep dive into the quirks, habits, and backstories that shape icons

Kevin Systrom turned ugly phone photos into global addiction.

He co-founded Instagram in 2010 after cutting Burbn down to its only useful part: photo sharing. His style was simple: remove what people ignored, keep what they used, and make the app easy to explain. Facebook bought Instagram for $1 billion in 2012, which is a strong argument for deleting things before adding more.

Cool Facts About Kevin Systrom

Five-Minute Start Rule: His procrastination trick is to work on the task for five minutes. Once he starts, he usually keeps going because starting is the hard part.

Smart Skimming: Systrom reads nonfiction with a goal in mind. He tries to find the useful ideas quickly instead of reading every page just to finish.

Fast Book Quitting: He has abandoned books after reading only a quarter of them if they were not useful. Kevin does not treat finishing a bad book as an achievement.

Learning by Outcome: When he wants to learn something, he frames it as “I want to be able to do X.” Then he looks for the book, person, or resource that gets him there fastest.

Simple Thing First: One of Instagram’s early working values was “do the simple thing first”. His view was that complexity often makes decisions look smarter while making the product worse.

Bottleneck Thinking: He applied the idea from The Goal that the slowest part of a system limits the whole system. At Instagram, the bottleneck was often executive decision-making, not engineering speed.

Decision Inventory: When he and Mike Krieger became bottlenecks, they created a document listing product decisions waiting on them. Then they held focused sessions where the only job was to decide.

Burbn Pruning: Instagram started as Burbn, an app with check-ins, plans, and photos. They cut almost everything and kept photos because that was the part people actually cared about.

Clear Product Story: Systrom wanted products people could explain in one sentence. If users cannot describe what it does, they also cannot spread it.

Watch-worthy clips

One video that got us thinking, and we think you'll like it too

This clip shows how to record audio in Apple Notes and get a transcript automatically. Simple feature, surprisingly useful.

Enjoying the newsletter? Please forward this issue to a friend who might enjoy it too 😊🙏🏻

It only takes 10 seconds. Making this one took us 9 hours…

If you are new here, what are you waiting for? ⬇️

Reply

Avatar

or to participate

Keep Reading