The tool becomes more powerful. Human skill becomes more valuable.
By John Wheeler
From Axes and Crosscut Saws
My great-grandfather came to America as an immigrant. He helped clear the great forests of the Pacific Northwest using axes and crosscut saws.
It was hard, skilled, physical work. A crosscut saw did not cut a tree by itself. The people using it had to understand the wood, the land, the weather, and each other. They had to know where the tree would fall and how to avoid being crushed beneath it. Strength mattered, but judgment mattered too.
Then the chainsaw arrived.
The chainsaw changed logging. It allowed one person to do work that had once required far more time and labor. Some tasks disappeared. Fewer people were needed to produce the same amount of lumber. The old way of working did not remain untouched.
But the chainsaw did not make knowledge worthless. It made a skilled worker more productive.
AI is the chainsaw of knowledge work.
AI Will Cause Real Displacement
I do not believe we help anyone by claiming that AI will not replace jobs. It will replace some tasks. It will change many jobs. It will probably remove some roles completely. Work that looked valuable five years ago may soon be cheap, fast, or automatic.
Software development is already showing us what this looks like. AI can now write a great deal of code. It can build features, find errors, write tests, and explain unfamiliar systems. Coding the way we did it five years ago is becoming obsolete.
That does not mean engineering is obsolete.
In fact, I think it puts a greater premium on engineering skill and creativity. When code becomes easier to produce, choosing the right problem becomes more important. System design becomes more important. Testing, security, business knowledge, and the ability to recognize a weak solution become more important.
The cost of producing code is falling. The value of knowing what should be built is rising.
This pattern will not stop with software. Routine writing, basic research, standard analysis, customer follow-up, scheduling, and paperwork will all change. Some work will disappear. Other work will be reorganized around people who know how to direct AI, test its output, and use it responsibly.
The Human Does Not Disappear
There is a popular story about a future in which AI replaces nearly everybody. I understand why that story gets attention. Fear travels quickly, and powerful technology makes dramatic predictions feel possible.
But capability is not the same as purpose.
AI may be able to perform a task, but it does not automatically know which task deserves to be performed. It can recommend an action, but that does not mean it has the authority to take it. It can produce an answer, but it cannot accept responsibility for what happens next in the same way a person or company must.
The future should not be built around removing people as if that were the goal. It should be built around giving people greater capacity while being honest about the workers who will need help adapting.
The chainsaw increased physical capacity. Excel increased our ability to calculate and analyze. The internet increased our access to information. AI increases our capacity to think, create, communicate, and act.
A more capable tool changes the work. It does not make the skilled person worthless.
M2M Is a Real Example
I am seeing this division of work firsthand while building Manufacturer to Menu, or M2M, an intelligent commercial management system for the foodservice industry.
I have spent decades operating companies and working with financial systems, business data, manufacturers, distributors, brokers, and sales organizations. That experience helps me define the product. I decide which business problems matter, what the system must know, which financial calculations must remain exact, and what an AI agent may or may not do.
AI performs much of the software development. It helps turn my direction into specifications, data models, tests, documentation, and working code. It can produce more in a short period than I could build alone using traditional methods.
Yet M2M would not become a useful product simply because an AI can write software. Without human direction, it could model the wrong commercial relationship, answer questions customers do not care about, use unreliable facts, or recommend actions that should require approval.
My role is not to type every line of code. My role is to define what should exist, explain the business, set the rules, and decide whether the result is useful.
AI supplies much of the production capacity. The human supplies direction and judgment.
This is not a claim that software jobs will remain unchanged. They will not. It is evidence that AI can move human work toward the parts of product development that require experience, creativity, judgment, and responsibility.
The Premium Moves Toward Skill
When a tool makes production easier, average output becomes less valuable. Better decisions become more valuable.
If everyone can generate a report, the advantage belongs to the person who asks the right question and knows whether the answer can be trusted. If everyone can generate code, the advantage belongs to the engineer who designs the right system and finds the failure that others missed. If every company can deploy an AI agent, the advantage belongs to the company that gives the agent reliable data, clear authority, good tests, and firm stopping rules.
We all have access to spreadsheets, but we do not all build the same financial model. AI will be similar. Access to the tool will become common. The advantage will come from how well people and companies learn to use it.
That will require more than prompt writing. It will require business knowledge, system thinking, communication, careful review, and the willingness to remain accountable for the result.
A Better Way to Prepare
Telling people that every job is safe is not honest. Telling them that humans will soon be useless is not helpful.
A better message is that the work is changing and we must change with it.
Workers should learn which parts of their jobs AI can perform well and which parts still depend on trust, physical presence, judgment, relationships, or deep experience. Companies should redesign work around the combined strengths of people and machines. Schools should teach students how to question, verify, design, and create, not merely how to repeat a routine process.
Leaders also have a duty to use AI with care. Faster production is not automatically better production. More output is not the same as more value. An agent that can take an action is not necessarily authorized to take it. Human responsibility does not vanish because software performed the task.
The goal should be to make people more capable while creating real paths for those whose work is displaced. That is harder than making predictions, but it is also far more useful.
The Tool Becomes More Powerful
My great-grandfather used axes and crosscut saws to help clear the forests of the Pacific Northwest. The tools changed, the work changed, and the number of people needed to perform some tasks changed too.
But knowledge still mattered. Skill still mattered. Judgment still mattered.
AI will produce the same kind of change in knowledge work, only faster and across far more industries. Some old methods will disappear. Some jobs will be lost. New roles will emerge, and existing roles will demand different skills.
We should face that disruption honestly. We should also refuse to confuse a more powerful tool with a less valuable human being.
AI can write much of the software. It cannot own the reason for building it.
The tool becomes more powerful. The human must become more skilled.
Get My Free Executive Wallpaper
The AI-Native Company – John Wheeler
Also read my views on:
The Autonomous Economy Is Coming – John Wheeler