AI isn’t confused — we are.
If you’ve ever typed a prompt into ChatGPT and thought, “That’s not what I meant,” you’ve met the mirror. AI amplifies whatever we give it — our thinking, our assumptions, our clarity — and it reflects them back, sometimes with stunning accuracy and sometimes in spectacular confusion. The difference isn’t in the model. It’s in us.
For most of my career, I’ve taught developers how to write changeable code — code that’s simple, extensible, and testable. But today, we’re not just writing code. We’re communicating with intelligence itself. And that demands a new kind of craftsmanship.
When I pair program with an AI, I’m struck by how it rewards clarity of thought more than cleverness of syntax. I can’t rely on intuition or habit. I have to describe what I want in concrete terms: the purpose, the shape, the trade-offs. I have to explain why before how. In other words, I have to think.
The New Skill: Thinking in Context
AI collaboration isn’t about mastering the perfect prompt template. It’s about learning to think in context. Before I ask an AI to do anything, I clarify four things — what I call the COAT framework:
1. Context: What situation are we in?
2. Objective: What outcome do I want?
3. Audience: Who is this for?
4. Task: What exactly needs to be done?
When those four dimensions are clear, the AI almost always delivers. When they’re fuzzy, it drifts — not because it’s “wrong,” but because I haven’t set the boundaries of meaning.
Developers have long known that ambiguity kills software. Now we’re discovering it kills prompts, too. The better we articulate intent, the better AI can collaborate.
From Code to Communication
When we say “AI will write code for us,” what we really mean is “AI will respond to our thinking.” The quality of the code depends on the quality of the conversation. That’s why developers who are great communicators — in naming, design, and explanation — are thriving in this new landscape.
We’re entering an era where code is no longer the only medium of creation. Conversation is.
The dialogue itself is the design document, the test plan, and sometimes even the implementation. This doesn’t diminish our role as programmers — it elevates it. Because to converse well with AI, we must understand our own thinking deeply.
The Return of Intentional Design
In the early days of Agile and Extreme Programming, we talked about intentional design — shaping systems by understanding their purpose. Then the industry sped up. We traded clarity for velocity, architecture for story points. Now, with AI, we’re being invited — or perhaps forced — to slow down and think again.
AI doesn’t reward busyness. It rewards clarity of intent.
When I tell it, “Generate a data model for an agile training system that can evolve as we add courses,” it listens for evolution, training, system.
When I say, “Write code for course management,” it gives me exactly what I asked for — and nothing I actually wanted.
Intentionality is the real superpower. Not prompting tricks, not secret hacks — just the discipline to know what you mean before you say it.
The Human Advantage
People often ask me, “Won’t AI make developers obsolete?” I smile, because the opposite is true. AI automates typing, not thinking. It frees us from mechanical labor so we can do what only humans can: model, abstract, perceive patterns, and care about meaning.
AI needs us to be more human, not less.
It needs our curiosity, our ethics, our empathy. It needs us to ask better questions. It’s not here to replace our creativity — it’s here to reveal how much of it we’ve been outsourcing to habit.
Clarity Is a Virtue
The best developers I know aren’t just technical. They’re clear thinkers. They name things well. They separate concerns. They write code that tells a story. Now we have a new collaborator — one who magnifies every ambiguity and every assumption. If we embrace that mirror, we’ll become not just better coders, but better communicators.
AI doesn’t need us to write more code.
It needs us to think more clearly, express more precisely, and design more intentionally.
Because in the end, clarity is the code — and that’s what AI understands best.
