Your boss just slammed a chemistry textbook on your desk. He demands you build a water molecule before 5 PM. He’s betting you can’t. He’s wrong.
This isn’t about chemistry. It’s about the dirty secret of every product manager who’s ever been cornered by a domain expert wielding jargon like a weapon. The secret? Domain expertise is a cage. Abstraction is the key that unlocks it.
Let me show you how a product manager and an engineer turned a chemical bond into a spreadsheet, and why that changes everything about how you design systems.
The Setup: A Boss Who Thinks He Knows Better
Zhang, the domain expert, was already humiliated. He’d bet his reputation on a physics problem and lost. But Li, the boss, is relentless. He walks in with a copy of Inorganic Chemistry, flips to a diagram of a molecule, and grins. “You guys cracked particle physics and e-commerce, sure. But chemistry? Chemistry is about shared electrons. Your engine doesn’t even have electrons. How can you simulate a molecule?”
He’s not asking a question. He’s issuing a death sentence. “By 6 PM, I want to see a water molecule—H₂O—synthesized in your engine. If you can’t make two hydrogens and an oxygen stick together, your engine is scrap metal.”
My engineer, Xiao Feng, is sweating. I pick up a whiteboard marker.
The Twist: You Don’t Need Electrons
I draw three circles. Two small ones labeled H, one big one labeled O. “Boss, you’re thinking about electrons. I’m thinking about boundaries. Western chemistry says a bond is electron cloud overlap. In system theory, a bond is a balance sheet settlement at the boundary of two structures.”
I write three words: Unit, Background, Boundary. Then I drop the formula that makes everything click:
ΔE = χ₁·χ₂·ΔL − ξ·χ₁·χ₂
“χ is the ‘bite’ of each atom—how hard it pushes. ΔL is the alignment between them. When they get close, the first term is repulsion. But the second term—the conduction term—dissolves that repulsion if the connection rate ξ is high enough. When ξ hits exactly 0.5365, the repulsion cancels to zero. That’s a covalent bond. It’s not magic. It’s accounting.”
Li frowns. “What about dynamic reactions? The Belousov-Zhabotinsky reaction oscillates between colors. Can your static accounting handle that?”
I smile. “That’s where the Five Elements come in. Water feeds wood, wood feeds fire, fire feeds earth, earth feeds metal, metal feeds water. That’s a limit cycle. The oscillation isn’t chemistry—it’s a lean manufacturing loop between production and consumption rates.”
The Proof: Numbers Don’t Lie
Xiao Feng runs the simulation. The charts hit the big screen.
Experiment 1: Water molecule synthesis — The boundary stress curve starts high, then drops. At ξ = 0.5365, stress hits zero. The atoms lock. A covalent bond without a single electron in sight.
Experiment 2: B-Z oscillation — The reactant and product concentrations oscillate in perfect sine waves. The engine never saw a reaction rate equation. It just followed the five-element loop. The result? A chemical reaction simulated by a product manager’s abstraction.
Li stares at the screen. Then he sends a voice message: “You… you really turned chemistry into mechanics? You didn’t even look at the electrons? Fine. I’m convinced.”
But he’s not done. “What about tower collapse? If you stack these blocks, how do you know it won’t fall?”
I grin. “That’s the next chapter. System aging and collapse prediction.”
The Lesson: Stop Letting Jargon Drive Your Architecture
Here’s the uncomfortable truth for every product manager reading this: You’ve been trained to respect domain expertise. But domain expertise is often just a fancy name for tunnel vision.
When a stakeholder says “we need high-concurrency handling,” they mean “make it fast.” When they say “shared electrons,” they mean “two things stick together.” Your job isn’t to learn their vocabulary. Your job is to find the universal operator—the χ and ξ—that collapses all their specialized requirements into a single configurable rule.
This isn’t just for chemical simulations. It’s for e-commerce, finance, logistics, anything. Every business problem is a flow of resources meeting a boundary constraint. Find the abstraction. Build the one engine that runs them all. And when your boss brings a textbook, don’t flinch. Hand him a whiteboard marker.
FAQ
Q: Isn't this just a toy model? Real chemistry is far more complex.
A: Yes, real chemistry has quantum mechanics, but the point isn't to replace chemistry—it's to show that the underlying structure of any system can be expressed in terms of boundary stress and equilibrium. The abstraction works for any domain where resources interact at boundaries, from atoms to e-commerce checkouts. The complexity is in the details, but the pattern is universal.
Q: How do I apply this to my product work without a whiteboard full of formulas?
A: Start by asking: what is the fundamental 'unit' of my system? What boundaries does it interact with? What are the 'stresses' (load, cost, time) and 'conductions' (connections, integrations)? Map every business requirement to a change in χ (bite) or ξ (connection rate). You'll find that 80% of features are just different configurations of the same two knobs.
Q: Doesn't this undermine the value of domain experts?
A: Domain experts are invaluable for understanding the 'what' and 'why' of a problem. But the 'how' of system architecture should be driven by universal abstractions, not domain-specific jargon. The best product managers respect expertise without being trapped by it. The contrarian take: if your system can't be explained with a half-dozen operators, you're overcomplicating it.