AI can generate a decent recipe from three random ingredients in about four seconds. What I care about is whether you can.
The game
Ask an AI for three or four random ingredients. Nothing curated, nothing safe, just whatever it picks. Then, before you ask it for a recipe, stop, close the chat, and try to come up with something yourself.
You aren’t trying to beat the AI. You’re using the constraint it hands you as a prompt for your own thinking.
Open-ended creativity is harder than constrained creativity; a blank canvas is more paralyzing than a limited palette. When you have four ingredients and have to make something work, your brain can’t drift. What flavour profiles are here? What textures? What technique could bridge these into something that makes sense on a plate? Is there a cuisine that naturally pairs two of them and could point toward the third?
That’s systems thinking as much as cooking: pattern matching across domains, recombining what you know, working under a constraint toward something new. It’s also the kind of thinking that makes you better at solving hard problems in code.
What it feels like
The first few times it’s uncomfortable. You’re used to recipes, which are instructions that remove the need to think. Constraints without instructions make you build the reasoning yourself, and the brain resists that at first because it’s more work.
After a few sessions something shifts. You start noticing that your first instinct, the obvious and safe combination, is just your fastest idea and rarely your best one. If you sit with the ingredients a little longer, a second and third layer of ideas shows up, and the interesting stuff is usually in that second layer.
I’ve had sessions where the AI’s recipe and mine were completely different and both would have worked. I’ve had sessions where the AI came up with something I’d never have thought of and I immediately wanted to make it. And sometimes I liked my idea better. I don’t keep score; the thinking is what I’m there for.
Why constraints help
Creativity research has a well-known finding that constraints improve output. Poets working in strict forms like sonnets and haiku often produce more inventive work than in free verse, because you can’t say the obvious thing in a fixed number of syllables with a set rhyme scheme. You find a less obvious thing that fits, and it’s often better.
Design, engineering, and cooking work the same way. With unlimited options there’s no pressure to be inventive, so the work tends to be mediocre. When something is ruled out, your brain goes looking for alternatives it wouldn’t otherwise have considered.
Random ingredients do this automatically. You didn’t choose miso paste, rhubarb, and lamb, so you can’t fall back on a known recipe.
It reminds me of the line from Iron Man: “Tony Stark was able to build this in a cave, with a box of scraps!” The cave is what forced him to be brilliant.
The neural angle
What you’re training is associative thinking, the ability to link ideas across domains that don’t obviously connect. It’s the same faculty behind good architecture decisions, good debugging instincts, and good design sense. I don’t think of it as a separate “creative” skill for artistic work; it’s a general cognitive ability.
Most of us let it atrophy. We live in a world of search, and whenever we hit a gap in knowledge the reflex is to look up the answer. That’s often the efficient call, but it means we rarely practice building the bridge ourselves from what we already know.
Cooking under constraints is low stakes. Nobody ships broken code because your miso-rhubarb experiment flopped. The muscle you use to figure it out is the same one you need when you’re staring at a hard problem nobody has written a Stack Overflow answer for yet.
How to run it
It’s deliberately minimal:
- Open your AI of choice and ask for four random ingredients. Give it a rough category if you like (e.g. “one protein, one vegetable, one pantry staple, one wildcard”) or let it pick freely.
- Close the chat before it says anything else.
- Spend five to ten minutes thinking. Write down ideas. Think in terms of flavour (sweet, salty, acid, bitter, umami), texture (crunchy, soft, fatty, light), and technique (raw, roasted, braised, fermented).
- Only after you’ve genuinely tried, open the chat and ask for a recipe using those ingredients.
- Compare, just to see where your thinking went versus where the AI’s went.
Sometimes imagination alone isn’t enough. Our mental model of how something tastes or smells is surprisingly weak in the abstract. If you’re stuck, get the ingredients out. Smell each one and hold them next to each other. Your nose picks up things your brain doesn’t consciously process, and a combination that sounds wrong on paper can make complete sense once you’re standing in the kitchen with both in your hands. The weirdest pairings often work for reasons you only find out by trying them.
You don’t have to cook anything, since the exercise is in the thinking. But if you do cook it, the meal tends to feel more satisfying than one you looked up, because you made it yours before you made it real.
The bigger point
AI is an extraordinary tool for generating answers. The risk of instant answers is that you stop practicing coming up with them yourself. Retrieval gets stronger while synthesis weakens.
Letting AI set the constraint and then stepping away before it solves the problem flips that around. The tool gives you the challenge, you do the work, and then you compare notes.
It’s a small habit. In my experience the people who stay sharp, who can reason through genuinely new problems without a template, keep finding ways to exercise that faculty. Cooking under random constraints is one of the stranger ways I’ve found to do it, and one of the most enjoyable.