2026-02-14

The Founder's Nonlinear Capability

In startup and investing contexts, "nonlinear capability" is mentioned a lot. It often sounds like talent, luck, or a trait reserved for "genius founders."

But in my view, nonlinear capability is not mysticism. It is a capability structure that can be deliberately forged.

More importantly, it does not come from working harder. It comes from understanding leverage, uncertainty, and how industry plus technology amplify each other.

1. Why linear capability often fails in startups

Most people are strong at linear capability:

• Put in more time, get proportionally more output
• Stabilize, refine, and standardize execution
• Optimize under clear goals and stable rules

These capabilities are highly valuable in mature systems: large companies, established industries, and clear paths.

But early-stage startups are not stable systems. They are high-uncertainty, path-dependent systems.

Startup failure is often not about effort. It is about working on things that cannot be amplified, even if executed perfectly.

2. What "founder nonlinear capability" means to me

For me, nonlinear capability can be broken down into several concrete parts.

1) Intuition for leverage

The first layer is not execution. It is judgment:

Which variables, once true, can amplify outcomes exponentially?

When I shifted from vehicle engineering to deep learning in 2019, it was not because I "liked coding more." It was because I judged that cars and physical systems were being restructured by software and AI.

That was not an incremental optimization. It was a directional bet.

Nonlinear capability often shows up in that choice, not in how hard you grind afterward.

2) Decision quality × amplifiability, not decision volume

Founders do not win by making more decisions. They win by making a few decisions at key nodes that can be amplified.

For example:

• Choosing the track, not polishing features
• Choosing the wedge, not expanding boundaries
• Choosing whether to keep validating or cut quickly

Many smart people fail not because they make poor decisions, but because they keep making good decisions in low-leverage zones.

3) Real nonlinearity comes from reversibility, not bravery theater

On the surface, nonlinear founders can look aggressive. In reality, they are highly sensitive to downside.

What they care about is:

• Whether failure is controllable
• Whether cost is recoverable
• Whether learning is transferable

Nonlinearity is not all-in gambling. It is using low-cost experiments to validate high-leverage hypotheses.

4) The ultimate form: self-reinforcing systems

Nonlinearity truly appears when a system starts to amplify itself, and the founder is no longer the growth bottleneck.

This can appear as:

• Product-led propagation
• Data/model flywheels
• Workflow embedding
• Network effects or scale effects

Before that, even a strong founder is still a linear amplifier.

3. Why smart and hardworking still does not equal nonlinear

This is easy to miss but critical:

• 10% technical improvement does not mean 10x business results
• More complex models do not mean stronger moats
• Harder problems do not automatically mean higher value

I have seen many people who are extremely smart and hardworking. Their issue is not capability. Their issue is that they keep solving linearly improvable problems.

In early startups, what is scarce is not "doing things better." It is doing the right things that can be amplified.

4. Why I use entrepreneurship to forge nonlinear capability

For me, entrepreneurship is a brutal but efficient training ground.

It forces me to face all at once:

• Uncertainty
• Real feedback
• Long-horizon game dynamics
• Organizational and human complexity

My current core product is essentially an experiment: can I turn personal judgment, technical understanding, and abstraction into a system the market can repeatedly validate and repeatedly amplify?

Closing: nonlinear capability is not a label, but a training outcome

Nonlinear capability is not a persona line, and not a slogan like "I want to do big things."

It is more like a filter:

If an effort has uncontrollable downside, no amplifiable upside, and no transferable learning, it is probably not a good nonlinear training ground.

For me, the meaning of entrepreneurship is not guaranteed success. It is whether I choose to place myself where high-leverage decisions keep happening.