Skip to main content

Manifesto

AI doesn't solve poorly defined problems faster. It builds them wrong, faster.

1. We're solving the wrong problem​

Everyone is racing to build with AI. Faster prompts. Cheaper tokens. Better code. The conversation is dominated by speed, cost, and technical wonder. And we get it — these things are measurable, demonstrable, impressive.

But they're the wrong metrics.

We've taken the oldest mistake in product development — building the wrong thing, fast — and given it a superpower.

2. The bottleneck was never the build​

It wasn't before AI. It isn't now.

The bottleneck has always been the same: understanding the problem deeply enough to propose a solution. Earning the right to build by doing the hard, slow, human work of listening, asking, and aligning.

AI didn't change that. It just made ignoring it more expensive.

3. Acceleration in the wrong direction​

Anti-pattern: Build-First Bias

A poorly defined problem, fed into an AI, produces a perfectly built incorrect answer. At scale. In minutes.

This is not progress. It's acceleration in the wrong direction.

4. What we believe​

  • The problem is sacred. The most valuable thing you can do before touching AI is to understand, precisely, what problem you're actually solving — not the problem you assume exists, not the easiest problem to solve, but the real one, found through genuine Discovery with the people who live it.

  • Alignment is not a meeting. It's a process of iteration, feedback, and organizational consensus that must be completed before writing a single line of context for an AI.

  • Context Engineering is the highest-leverage skill of the AI era. The quality of what comes out is a direct function of the quality of what goes in. This is not a technical skill. It's a thinking skill.

  • Building trends toward zero cost. Once the context is precise, the build is not the product. The thinking is the product.

  • The market closes the loop. The market is not the end of the process. It's the beginning of the next iteration. Real-world usage signals feed back into the context, close the loop, and make the system smarter over time.

5. Problem-Driven AI is pro-thinking​

tip

We're not asking you to slow down. We're asking you to aim before you fire.

The teams that will win the AI era are not the ones that build fastest. They're the ones that think most clearly, define most precisely, and align most deeply — before opening a prompt.

The problem is still the product. AI just changed what happens after you find it.