Blind spot 1: Building before you know the user
Many teams start writing code as soon as a requirement arrives, without understanding who the real user is or what problem they need solved.
Case: an AI customer-support bot
A team spent three months building a conversational support bot. After launch, they discovered that users actually needed a structured form—not an open-ended chatbot.
Why does this happen?
In the age of AI, tools such as and Cursor make software dramatically easier to produce. That makes more important, not less.
As execution cost falls, decision cost rises:
- Before: dominated; making anything was expensive
- Now: AI compresses execution, while choosing the right direction carries more of the risk
The problem has shifted from having no viable approach to having too many. AI can produce ten plausible designs in moments, but it cannot decide which one fits your users, constraints, or consequences. More options make judgment more important, not less.
Execution ability determines whether you can build something. In the age of AI, decision ability increasingly determines whether you build the right thing—and whether you build it well.
The rigour has not disappeared because AI can write code, documentation, or automation. It has moved upstream: define the problem clearly, make the decision deliberately, and verify the outcome. The lower execution cost gets, the more valuable that clarity becomes.