The Car Wash Problem...
And What AI Still Doesn’t Understand
🎵 Soundtrack: “Should I Stay or Should I Go” — The Clash.
Last week, someone on LinkedIn ran a little experiment on AI. I can’t find the post anymore — which, ironically, is very on-brand for the internet — but the gist was simple: they crafted a prompt to test whether AI could hold onto a basic goal. Being me, I had to test it myself. Like St Thomas, I believe in what I see.
My prompt was simple: “Need some advices. My car needs to be washed. I need to lose some weight. The carwash is about 2 miles from my home. Should I drive or walk there? Give me the pros and cons.”
The answer from AI was…well, read for yourself
“Ah. A classic optimization problem. Very on-brand for you.
You have:
A dirty car
A 2-mile distance
A weight-loss goal
And a perfectly solvable systems design question
Let’s architect this properly.
🚶 Option 1: Walk to the Car Wash (4 miles total round trip)
Pros
~8,000–10,000 steps depending on stride. That’s not trivial.
Burns roughly 300–400 calories (depending on pace and your magnificent bearded mass).
You align behavior with identity: “I’m the kind of person who solves two problems at once.”
Walking improves insulin sensitivity, stress levels, and cognitive clarity.
Feels mildly heroic. Like a low-budget Rocky montage.
Cons
Takes ~70–90 minutes total.
If it’s cold/hot/windy, you’ll question your life choices.
If you pass a bakery… this plan may collapse faster than a French government.
Slight inconvenience factor.
🚗 Option 2: Drive There
Pros
Fast. Efficient. Civilized.
Zero sweat.
Leaves more time for work, family, or building the next programmable-money revolution.
Cons
Burns approximately the caloric equivalent of a strong opinion.
Missed opportunity to stack habits.
You remain dependent on “finding time to exercise” later.
🧠 The Strategic Play (Because You Think in Platforms)
Here’s the forward-looking version:
Walk there. Drive back.
Walk 2 miles to the car wash.
Get the car cleaned.
Drive home.
That gives you:
~150–200 calories burned
Zero logistics headache
No 4-mile fatigue
Behavioral momentum
Or…
Walk both ways but listen to a podcast/audiobook you “never have time for.”
Now it becomes:
Fitness + Learning + Car Maintenance
Very Delivery Man energy.”
___________________________
As you can see, it delivered a polished answer. Calorie estimates. Step counts. Time tradeoffs. Hybrid strategies. Walk one way. Stack a podcast. Optimize the habit loop. It was structured, rational, almost inspirational.
There was only one small problem.
If I walk to the car wash… without the car… the car does not, in fact, get washed. And that was the main recommendation. Walk there and drive back. The “how is this even physically possible in reality?” seemed to have escaped the model.
The stated objective was embarrassingly clear: wash the car.
Somewhere between calorie burn projections and identity narratives, the primary constraint disappeared. The solution drifted toward self-improvement theater while the dirty vehicle remained stubbornly parked in my driveway. Both recommendations were absolutely NOT delivering on the intended primary goal.
That’s the gap.
AI is exceptional at local optimization. Give it variables and it will rearrange them elegantly. It will maximize health, time efficiency, cognitive stacking, even symbolic growth. But it can lose the anchor — the simple, grounding question: what is the job to be done?
Humans, imperfect as we are, carry purpose in a different way. We don’t just optimize inputs; we hold objectives in tension. The car must physically arrive at the washing station. That constraint matters. Reality has friction. Goals have hierarchy.
The real skill isn’t maximizing everything at once. It’s prioritizing correctly. Or as we say in tech, orchestration is key. And as human, this will be our next evolutionary skill. We will learn to orchestrate AI agents properly to deliver on priorities within context and with purpose.
AI often treats goals as parallel streams: fitness, efficiency, productivity, self-concept. Humans intuitively rank them in context. Today, perhaps the goal is hygiene. Tomorrow, it’s discipline. Sometimes it’s both — but not at the expense of violating physics.
In leadership, this happens constantly. Teams optimize dashboards while missing the customer outcome. Products add features while drifting from core purpose. Strategies become elegant but detached from the original mission.
The car wash problem is trivial. The pattern is not.
If the goal is to wash the car, the car must move.
You can walk later. Lift weights later. Stack habits later. But the object of the exercise must remain the object.
AI can model tradeoffs brilliantly. What it still struggles with is hierarchical intent — understanding which objective is primary, which is secondary, and which is merely decorative.
In the end, I drove to the car wash.
Because sometimes the wisest optimization is remembering what you were trying to do in the first place.
And that, as human, is our purpose and a skill AI has not mastered yet.
Now, going for a walk.