Second-Level Thinking: Why Almost Every Builder Makes the Wrong Decisions (And Doesn’t Even Know It)
On Monday I covered Munger and incentives. On Saturday, Naval and permissionless leverage.
Today I close the trilogy with the thinker almost nobody in the tech world mentions: Howard Marks.
And it makes sense that we ignore him. He’s a fund manager. He talks about high-yield bonds. Nothing to do with Next.js or launching a SaaS.
Or so I thought.
The Problem With How Most of Us Think
There’s a question Marks has been asking for decades—one that stings a little if you’re honest with yourself:
Why do you think your analysis is right if you have access to the same information as everyone else?
In investing, this destroys portfolios. In product, it destroys startups.
Marks calls it first-level thinking: the ability to reach an obvious conclusion from obvious information. Everyone can do it. And that’s exactly the problem.
“First-level thinking is simplistic and superficial, and just about everyone can do it. Second-level thinking is deep, complex and convoluted.”
— Howard Marks
First-level thinking says: “AI is growing massively, I’ll launch an AI tool.”
Second-level thinking says: “Everyone wants to launch an AI tool. What does that mean for market saturation? Where’s the opportunity others are ignoring precisely because they’re all chasing the obvious thing?”
It’s not about being contrarian for the sake of it. It’s about thinking one step further.
The Real Difference Between the Two Levels
This took me a while to internalize because it sounds like armchair philosophy.
The concrete difference is this:
First level: What’s happening?
Second level: What does everyone think is happening, and what does that imply about the reality nobody’s seeing?
In 2026, after last year’s generative AI hype peak, many builders are making the same first-level move: “AI isn’t novel anymore, I’ll find a different niche.”
But second-level thinking asks: What happens when the majority abandons a market for being saturated? Who stays? What problems remain unsolved precisely because everyone went after the same obvious ones?
And here’s the interesting part: you don’t need to be smarter than everyone else. You need to think one level higher.
The Pendulum and Tech Hype Cycles
The second Marks framework I apply most is the market pendulum.
His observation: investor psychology swings like a pendulum between excessive optimism and excessive pessimism. It almost never rests at the midpoint.
“The mood swings of the securities markets resemble the movement of a pendulum… spending very little time at the midpoint.”
— Howard Marks
This isn’t just about the stock market. It’s about how collective attention moves toward any technology, niche, or idea.
I watched this happen with no-code tools in 2024, with AI agents in 2025, and I’m watching it again in 2026 with certain infrastructure stacks. The pendulum always swings. The question is: at what point in the swing are you making your decisions?
When everyone is euphoric about something → first level says “jump on the bandwagon.” Second level asks “what price am I paying for this euphoria?”
When everyone is abandoning something → first level says “it’s over.” Second level asks “is there real opportunity here beneath the panic?”
For a builder, this translates into something very concrete: perceived saturation and real saturation are not the same thing.
How I Apply This to Product Decisions
Three second-level questions I use when evaluating an idea or deciding whether to pivot:
1. What is everyone assuming to be true?
Most “saturated” markets are only saturated in the obvious segment. Second-level thinking looks for the segment where the common assumption doesn’t apply.
2. If my analysis is correct, what would have to be true that contradicts what the market believes?
This is the uncomfortable question. If my conclusion is right and most people are wrong, then there’s something the market isn’t seeing. What is it? Can I defend it with solid reasoning or just gut feel?
3. At what point of the pendulum am I making this decision?
If I launch at peak euphoria, I pay the maximum price in time, attention, and competition. If I launch at peak pessimism, I have an entry window. Marks doesn’t predict when the pendulum changes direction, but he teaches you to recognize the extremes.
What Nobody Says About Thinking Differently
There’s a critical nuance Marks emphasizes that most people omit when talking about “being contrarian”:
It’s not enough to think differently. You have to think differently AND better.
Being contrarian by principle is just as superficial as following the crowd. The edge isn’t in going against consensus—it’s in having a more rigorous analysis than consensus.
Many builders in 2026 are being contrarian about AI: “all that is hype, I’ll build something traditional.” That’s first-level thinking dressed up as second-level. Real second-level asks: within AI, which part of the hype has genuine foundation and which is pure narrative? Where is the value that will survive the cycle?
Closing the Trilogy
Munger teaches you to think in incentives and mental models.
Naval teaches you where to focus your energy and leverage.
Marks teaches you how to evaluate every decision without being swept along by what the majority thinks.
All three share something: the discipline to question the obvious analysis before acting.
In product, in code, in business, the most valuable decisions are rarely the most obvious ones. They’re the ones that require one extra thinking step most people don’t take.
Your Action This Week
Take the next major pending decision—launching something, pivoting, entering a niche, abandoning a project—and ask yourself these two questions before acting:
- What is everyone assuming to be true in this context?
- What does that imply about what nobody is looking at?
You don’t need a perfect answer. You need the habit of asking the question.
That’s what separates first level from second level.
