I didn't expect food to be part of this series.
I was wrong.
The Hidden Tax
Decision fatigue is a specific, documented cognitive phenomenon — not a metaphor. Roy Baumeister's research established that self-control and executive decision-making draw from a shared resource that depletes across the day. You start with a full budget. Every decision you make — trivial or significant — withdraws from it. By the time you've made a dozen decisions, the quality of each subsequent one degrades measurably.
What the research doesn't make obvious is just how many decisions happen before you sit down to work.
What to eat for breakfast. Whether there's enough of it. Whether to go get something. What to get. What that will cost in terms of time. Whether you should have prepped something last night. What you'll eat for lunch. Whether you need to stop somewhere. What's in the fridge. Whether that's still good. Whether you should order something. What's close. What's worth it.
That's not one decision. That's a tax bracket.
It compounds. By the time some people hit their desk at 9am, they've already made 15-20 micro-decisions about food, clothing, and logistics. Research from the Journal of Consumer Psychology suggests we make roughly 35,000 decisions a day — and the cognitive cost doesn't stay proportional to the decision's importance. The brain doesn't know that "what's for lunch" matters less than "should I restructure this client engagement." It just processes. And it gets tired.
Where I Was Before
I used to handle food the way most people handle it: reactively.
Breakfast was whatever was fast. Lunch was whatever was close or whatever had the shortest decision path when I was already hungry. Dinner was 30 minutes of half-hearted browsing — recipe sites with 600-word origin stories before the ingredients, three open tabs, a pivot to something simpler, another pivot, eventually just ordering something because the friction had fully defeated me.
The time cost was real. The attention cost was worse.
By mid-afternoon I was making food decisions from a depleted state, which meant the decisions were worse, which meant I was less satisfied, which meant I thought about it more. The decision wasn't closing. It was creating a low-grade background process — a cognitive tab I didn't open intentionally and couldn't fully close.
That's not a nutrition problem. That's an architecture problem.
The Structural Fix
I moved food decisions to once a week.
One prompt to Claude, Sunday evening, from a full cognitive state. The prompt structure is simple:
Here's what's in the fridge. Here are my constraints this week. Build me a meal plan for 5 days with a consolidated shopping list.
What comes back is complete: breakfast, lunch, dinner, snacks, one list organized by store section. The plan reflects what I actually have, what I actually want, and what I can actually execute given the week's calendar.
The shopping takes 25 minutes because I'm executing a list, not making decisions in the aisle. The meals happen because the plan exists and the ingredients are already there.
The key architectural shift: food decisions now happen once, from a rested state, with full context — not five times a day, from a depleted state, with no context except hunger.
What This Cost Me
Nothing. That's the point.
The weekly planning session takes about 10 minutes. That replaces 30-45 minutes of reactive browsing every day I'd otherwise spend figuring it out. The math isn't even close. But the time savings are secondary to what I'm actually recovering.
The first hour after lunch used to be my worst cognitive window. I'd made too many decisions by that point — work decisions in the morning, food decisions at noon — and I'd arrive at the afternoon depleted. Moving food decisions out of that daily cycle means I arrive at afternoon work with more budget intact.
The background process closes. The open tab shuts. The cognitive overhead of "what am I eating" is no longer running in parallel with everything else.
One Decision, Many Downstream Wins
This is what I mean when I say architecture. A single structural change — where does the food decision happen, and when — produces downstream improvements that have nothing to do with food.
Better afternoon focus. Less friction at dinner. No 6pm decision spiral. Shopping trips that take 25 minutes instead of 45. Less food waste. Less ordering out from exhaustion.
None of those outcomes required willpower. None required a new habit built through discipline. They required one design decision upstream.
Obama eliminated his clothing decisions — two suit colors, end of deliberation — not because he was lazy but because he understood the budget. Every domain of life that generates high-frequency, low-value decisions is a candidate for the same treatment.
Food was the easiest one I hadn't solved yet.
The Prompt I Actually Use
Not abstractly. Literally:
It's Sunday. I have: [list]. I'm avoiding: [constraints]. My week looks like [calendar context]. Build a meal plan with shopping list.
That's it. Claude returns structure. I execute against it. The week's food decisions are done.
No recipe rabbit holes. No Pinterest boards I'll never revisit. No 11pm realizations that tomorrow's plan requires an ingredient I don't have.
The decision is made once, from the right cognitive state, with the right context — and then it's not a decision anymore. It's just what's for dinner.
What's the highest-frequency, lowest-value decision you make every day that you've never thought to systematize?
This is part of The Layer — a series on building AI as the operating system beneath your decisions. Read Episode 0: The Layer.
