The Machine Didn't Take Your Craft. You Gave It Up.
Written by Derek Chua, digital marketing consultant and founder of Magnified Technologies. He runs a multi-agent AI system in production and thinks regularly about what humans should actually be doing in an AI-assisted world.
There is an essay circulating right now that is making developers uncomfortable. Not because it is wrong, but because it is direct.
It starts with a woodworker. New machines arrive. They are faster, more precise, cheaper. The market adjusts. The woodworker sets down his chisel. And we are told something tragic has happened.
Developer David Abram, writing on his personal blog, is not buying it.
Key Takeaway: Giving an AI tool the cognitive load you were meant to carry is the real risk of AI adoption, not the tool's existence. The businesses and employees who thrive with AI are those who use it to amplify their thinking, not replace it. That distinction determines whether you become more valuable or less.
The Argument That Is Spreading
Abram's piece is titled "The machine didn't take your craft. You gave it up." It surfaced on Simon Willison's weblog this week and has been circulating steadily since.
The argument, stripped down: if you love your craft, no tool can take it from you. The woodworker who sets aside his chisel the moment a machine exists did not love the craft. He loved the position the craft gave him. The developer who says "AI is taking away my coding" is not describing an external force. They are describing a tension between what they enjoy and what the market now rewards. That tension is real. But it is not new. And it is not the machine's fault.
The sharper point comes when Abram, speaking as a working developer, says:
"The hardest parts of the job were never about typing out code. I have always struggled most with understanding systems, debugging things that made no sense, designing architectures that wouldn't collapse under heavy load, and making decisions that would save months of pain later. None of these problems can be solved by LLMs."
He is not cheerleading for AI. He is pointing at something most AI hype cycles miss: the real work was never what the tool now handles.
Why This Matters Beyond Software Development
I keep this framing in mind when talking to business owners and team leads about AI adoption.
The conversation often goes like this: someone wants to use AI to "handle content" or "take care of customer queries" or "automate reports." They are thinking about tasks. What they should be thinking about is judgment.
Tasks are the visible surface of most jobs. Judgment is what the job is actually for.
An account manager's value is not in sending status updates. It is in reading the client relationship, knowing when to push back, spotting the project risk before it becomes a crisis. A marketer's value is not in writing copy. It is in understanding why this audience responds to this message in this context.
Give an AI tool your tasks. That is right. Keep your judgment. If you let the tool carry the judgment too, you have not become more efficient. You have become less relevant.
The risk Abram identifies is what he calls "the abdication of reason from within." That is the trap. Not the existence of the tool. The decision to stop thinking.
At Magnified, I have built and run AI agents for content creation, client work, and research. The agents handle volume and speed. I handle quality gates, strategic direction, and the calls that require actual contextual judgment. That balance is not accidental. It is the design.
What This Means If You Are Building Your Team Around AI
If you are a business owner or team leader, this framing has practical implications.
Automate the task, protect the thinking. When you use AI to handle customer queries, email drafts, data summaries, or content briefs, the human on the other side should be doing more judgment, not less. If AI is automating the easy parts and the human is just approving the output without thinking, you have replaced skill with rubber-stamping. That is a fragile position.
The employees who thrive will be the ones who use AI to go deeper, not shallower. The developer who uses AI to handle boilerplate writes more complex systems. The marketer who uses AI to generate options develops sharper taste. The analyst who uses AI for data extraction asks better questions of the data. This is the path to becoming more valuable, not less.
Hiring should test for judgment, not just output. If your hiring process involves tasks that AI can now complete in 20 seconds, you are testing for the wrong thing. Test for the thinking that happens before the task. The framing of the problem. The decision about which task to do in the first place.
And if you are an employee worried about AI: Abram's point applies directly to you. If you reduce your professional identity to "the person who does this repeatable task," then you are right to feel exposed. But that reduction is something you are doing, not something AI is doing to you. The craft is the thinking. Own that part, and the tool becomes leverage.
What "AI + Humans" Actually Means
I use the phrase "AI + humans > AI alone" a lot, and it is not just a slogan. It is the practical reality I see in how these systems work in production.
AI alone produces volume without judgment. Humans alone have judgment but are limited by time and bandwidth. The combination, when done correctly, gets you volume and judgment. When done poorly, you just get volume. A lot of it, fast, without the thing that made it worth doing.
The woodworker's mistake, in Abram's framing, was not picking up the machine. It was putting down the thinking that made him a woodworker in the first place. One of those was a choice. The other was the craft.
We are at an early point in AI adoption where many businesses are still figuring out which parts of the work they are outsourcing to the tool. The ones getting it right are the ones who are very clear about what thinking they are keeping for themselves.
Frequently Asked Questions
Is AI actually replacing skilled workers, or is this concern overblown? Both are partly true. AI is replacing some task-based work, particularly anything repetitive and well-defined. But the concern that AI is eliminating the need for human expertise tends to conflate tasks with judgment. Most roles involve a mix of both. The task-heavy parts are increasingly automatable. The judgment-heavy parts are not, and are becoming more valuable as a result.
How do I know which parts of my job I should protect versus delegate to AI? A useful test: if the output requires you to understand the context, the stakeholder, the risk, or the constraint before you can evaluate it, that is judgment. Protect it. If the output can be evaluated against a checklist or standard format without contextual knowledge, that is a task. Delegate it.
Does this mean businesses should limit how much employees use AI? Not exactly. The goal is not to limit AI use but to be deliberate about cognitive load. If an employee uses AI to do their thinking for them, their own capability tends to atrophy over time. If they use AI to do more of the work that does not require their specific thinking, they tend to develop faster. The difference is whether the human is still making the important calls.
What should businesses do now to avoid the "abdication of reason" trap? Build review processes that require humans to articulate why they approved something, not just that they did. Make sure AI-generated outputs are evaluated against a standard the human understands, not just a "looks right" intuition. And train teams on what the AI is actually doing, so they can catch it when it is wrong. The trap closes when humans stop engaging critically with AI outputs. The fix is designing for engagement, not just efficiency.