johnmaconline chain logo johnmaconline / The250
Daily one-pagers about AI, leadership, technology, and being human.

johnmaconline

I'm writing to think, learn, and remember in public. I'll be here everyday.

August 6, 2026 2 minute read

Somebody Who Cares

HumanLeadershipThe250

We found ourselves in a situation.

Rented an apartment in Philly. Showed up to move in. It was the wrong apartment.

So let me clarify that. We signed the lease on the apartment that they gave us, but that was NOT the apartment we had been shown nor the one we wanted.

It was bait and switch, but (probably) an honest mistake, not malicious.

Why did this happen?

Because the rental agent dropped the ball. Didn’t do her job. Provided the paperwork for the wrong apartment. Didn’t check. Didn’t ensure. We trusted, and had no choice but to trust given the situation. Because how would we know?

So here we are. Full truck parked at the loading dock on Ludlow St with the flashers on.

First call was to the rental agent.

“No, that’s the apartment you wanted.” Then, “But you said…” Then, “I have in my notes…”

All wrong. All BS. All verifiably incorrect via the text message string. All her trying to push the blame onto us.

Then finally, “Well, that’s the lease you signed.” Unfortunately, the only truth of the conversation.

All we needed — all we wanted — was somebody who cared and would try to help. We weren’t moving into this apartment. So help us figure this out.

She had an opportunity here. She had an opportunity to help. She had an opportunity to advocate. She had an opportunity for true customer service.

She had an opportunity to care.

But she didn’t.

If you’re in the business of customer service, it’s gonna pay to be the person who cares about your customers.

August 5, 2026 1 minute read

Asking Questions

AIHumanTechnology

“Do AI models ask questions?”

ChatGPT:

“Not usually in the way a person asks questions.

During standard training, the model is mostly given examples and learns to predict the next token. It does not pause, notice confusion, and ask the trainer for clarification. The training system selects the data, calculates the error, and adjusts the model automatically.

Questions can appear in the training data, so the model learns how and when people ask them. Later stages may also specifically reward good clarification questions through human feedback or reinforcement learning.

Some newer training approaches let models generate questions for themselves, create practice problems, critique their own answers, or request information from tools. But that is a designed training loop—not spontaneous curiosity.”

Claude and ChatGPT don’t ask questions. They aren’t curious. They don’t care.

You ask questions. You are curious. You care.

Keep asking questions.

August 4, 2026 1 minute read

You’re a Cliche

HumanLeadershipThe250

I’m sorry. So am I.

I’m one of the hapless new homeowners from the Dr. Rick commercials who is “turning into my parents.” Defense does win championships. Proper tool selection helps. I like to be at the airport early.

I’m ok with it. Most times. I just hope the way my youngest son laughs at me is some form of love.

I’ve just spent a few days vacationing in a planned community near the coast. This place is so full of clichés, you could make a game out of it. Point randomly and guess the back story. You’d probably be pretty close.

We’re all a cliché in some way or another. When others describe you, they use it. Even if they say “unique,” that’s a cliché.

You change it, of course.

Start a new practice. Change a current one. Take a different road. Do the hard thing. Make a plan. Change the plan. Go the other way. Do more. Do less.

However, rather than fighting a cliché, fight the cliché.

The one you don’t want to be.

August 3, 2026 1 minute read

Compromise

HumanLeadershipThe250

Compromise is the art of meeting in the middle, so everybody is somewhat happy.

But also somewhat not happy.

Compromise is both a purposeful strategy and a lowest common denominator when we don’t know what else to do.

Relationships, leadership, and community building all require some degree of compromise to be successful. But also, you can’t be successful if everything’s a compromise.

Compromise is the art…

August 2, 2026 1 minute read

Non-Negotiables

HumanThe250

When you’re looking for a new job, you probably have some non-negotiables.

Maybe it’s location, pay, vacation time, coworkers, or mission.

I’m now convinced it’s also token budget.

You gotta ask.

What’s my token budget?

It’s a non-negotiable.

August 1, 2026 1 minute read

Plus

AITechnologyThe250

AI enables you personally to do more.

AI enables your organization to do more.

So do more.

More people. More ideas. More value.

Addition. Not subtraction.

July 31, 2026 2 minute read

Flexibility, Efficiency, and Optimization

HumanTechnologyThe250

You usually can’t have them all.

Most comp sci majors today learn how to program with languages like Java and Python. These are great, extremely flexible languages. They run on any processor and any operating system. The programmer doesn’t need to know anything about the hardware.

Java’s like a luxury SUV. You get in, turn it on, and drive. It goes anywhere and you already know how to operate it.

Most embedded code is written in C and C++. These are still somewhat flexible languages, but since you must compile them for a specific processor within a specific system, you must understand the hardware on which it is meant to run.

C/C++ is like a sports car with a manual transmission. It still goes anywhere, but requires some skill on your part to work well.

If you want the most efficient code, you write it in assembly language. Assembly code is specific to the processor, its peripherals, and its physical memory. It’s not portable to a different system.

Assembly is like a Formula 1 race car. It only drives on racetrack, and you definitely need skills to drive it.

If you need even better performance, you can optimize your assembly code. Optimization, though, requires deep knowledge of both the hardware and the inputs you’re giving it. The optimized code has zero flexibility. If you change one of the inputs, you will have to rewrite the code.

Optimized assembly is like a rocket car built to break the land speed record on the salt flats. It drives in one specific place, requires highly specific skills, and can break at any moment for a million little reasons.

The Java code might take an hour to develop.
The C code might take half a day.
The assembly code maybe all day.
Optimized assembly might take 2 weeks.

Don’t build a rocket car to drive to the grocery store. Also, you’ll never hit the land speed record in a cookie-cutter luxury SUV.

What are you trying to build? Who’s it for? What matters?

July 30, 2026 1 minute read

End of Your Career

HumanLeadershipThe250

Lebron’s coming to Philly to finish his career.

Didn’t sign for a large amount of money. Didn’t go back to his place of origin. Didn’t go back to the franchise he liked the best. Didn’t go back to the place his family is.

He’s finishing it on his terms.

I’m not at the end of my career, but I’m starting to think about it. How would I like to finish it? What are my terms?

It’s hard to finish on your terms if you don’t know what they are.

July 29, 2026 1 minute read

Next Version

AIHumanTechnology

Version 3 is the good one.

But what about 3.1 or 4?

I’m sure you’ve opened an app or a web page and thought to yourself, “What happened here? It’s all different!”

You may have even thought, “Damn it! Why do they keep changing things? It worked exactly the way I needed it to already.”

And, of course, now, they’re adding AI to everything. Whether you want it or not.

The next versions after 3 aren’t about you.

It might be about their competitors. It might be about their bottom line. It might be about vanity or status.

But the next version isn’t about you.

July 28, 2026 2 minute read

Commoditization of How

AITechnologyThe250

If you’re somebody with ideas, it’s a great time to be alive. It’s a builder’s paradise.

As long as you have access to tokens. But a reasonably small investment ($20-$200 / month) gets most hobbyists and many professionals plenty of tokens.

AI can and will build anything for you in the online, mobile, and virtual worlds. Just tell it what you want, and it will do it for you.

AI can and will show you how to build anything in the physical world. Just tell it what you’re trying to do, and it will show you how to do it.

Until recently, the value was in the how.

How do you build a paver patio? Hire a contractor who knows how.
How do you manage your finances? Hire a planner who knows how.
How do you build an app? Hire a software developer who knows how.

The how was scarce. How took learning and experience.

AI makes the how abundant. AI commoditizes the how.

But it doesn’t know the what or the who.

What do you want to build and for whom?

AI gives you the workshop. It’s up to you to figure out what you want to build.

Pin It on Pinterest