Coding Faster Isn't Enough

Two shifts are happening right now. They happen in parallel. And one makes the other harder.

What Getting Things Done Looks Like Now

AI is taking over execution at an increasingly rapid pace. Not always, not everywhere, but the trend is clear. Code that previously took days gets written in hours. Workflows that required three specialists can be deployed by one person with the right methodological understanding. This is no longer hype -- it's a fact beginning to show up in how organisations actually staff and deliver.

That changes what's required. The role shifts from producing every detail yourself to formulating the problem, steering direction, and taking accountability for the result. The bottleneck is no longer execution. The bottleneck is judgement.

But that creates a dual competency shift for anyone who has previously only coded.

The First Shift: Understand the Business

AI cannot guess why a system needs to be built. It cannot determine which problem is actually worth solving, which outcome is meaningful to the client, or whether an elegant technical solution addresses the right thing. That requires domain understanding, consequence thinking, and experience of what actually works in an organisation's daily reality.

Someone who has spent eight years coding solutions someone else defined isn't lacking competence. But that competence is something different -- and that shift takes time.

The Second Shift: Using AI for Execution

In parallel with understanding the business better, the same person needs to learn to let AI do what they previously did by hand. That is not a trivial change. It's an identity shift. Craft built up over years gets redefined. It requires new methodology, new ways of validating, new ways of owning the result.

These two shifts are therefore not happening in sequence. They're happening at the same time.

Why It's Hard in a Way Nobody Talks About

There's a third factor that makes all of it harder, and it rarely gets mentioned: the business model.

For consultants and freelancers who sell their time, speed is actively counterproductive. More billable hours means more revenue. The person who is fast with AI earns less by definition, if nothing else changes. Embracing AI without solving the business model is not rational -- it's an economic sacrifice.

That creates a paradox. The hourly billing model rewards three things: being slow, executing rather than understanding, and staying within what you already know. All three are exactly what the AI era requires you to stop doing.

Telling a consultant they "should embrace AI" without addressing the underlying incentive structure is not advice. It's preaching.

Who Is Actually Moving Fastest

What's interesting is who is moving fastest right now. It's rarely the coder who has learned to understand the business better.

It's the entrepreneur who has always understood the business, but until recently depended on others to execute. The person who knew exactly which problem needed solving, why it was valuable, and what the client was actually paying for -- but lacked technical execution capability and had to rely on intermediaries.

That dependency is now largely gone.

And that's an important signal. Those moving fastest into the AI era are not necessarily the most technically skilled. They're the ones who always had the answer to "why" -- and can now finally answer "how" without having to ask someone else.

That fundamentally confirms what AI actually amplifies: domain knowledge. Not technical dexterity.

What Is Actually Required

Coding faster isn't enough. Understanding the business isn't enough. And selling differently isn't enough.

All three need to happen. The competency shift -- understanding the business and using AI for execution -- is genuinely hard and takes time. The business model shift -- stopping selling hours and starting to sell delivered outcomes and understanding -- requires redefining what you are and what you offer.

It's an identity shift, not a tool change.

The one who pulls it off has a position that won't be competed away by the next model version. The one who just codes faster hasn't solved the problem. They've just shortened the time it takes.

See also: The Impressive and the Billable Are Not the Same Thing (series 28) and It's Not the Speed That Kills You (series 26)