Is AI a tool—or a replacer?
The concern is understandable.
AI systems can already analyze enormous quantities of data, generate documents and computer code, identify patterns, assist with engineering decisions and automate work that once required considerable human effort.
As these capabilities improve, workers naturally wonder whether AI will make them more productive—or eventually make their jobs unnecessary.
But perhaps we are asking the wrong question.
The history of industrial progress is largely the history of humans developing better tools and then discovering entirely new things they can accomplish with them.
The steam engine changed production. Electricity transformed the factory. Machine tools increased precision. Computers changed engineering and administration. CAD/CAM transformed product development. CNC machines changed machining. Industrial robots automated repetitive production.
Artificial intelligence may be the next major step in that progression.
The real question may therefore be:
Will we use AI primarily to replace human capability—or to multiply it?
In an August 2026 Times article, economist Paul Johnson argues that innovation is fundamental to long-term human and economic progress.
That argument has particular relevance to manufacturing.
Manufacturing has never been a static industry.
Every generation of manufacturers has had to adapt to technologies that changed how products were designed, produced, inspected and distributed.
Those changes frequently created anxiety about employment.
When machines became more capable, people understandably asked what would happen to workers performing the existing tasks.
Yet technological progress also created new industries, new occupations, higher productivity and capabilities that previous generations could scarcely imagine.
This doesn’t mean technological disruption is painless.
It means that evaluating a new technology solely by asking which existing tasks it can eliminate gives us only half of the picture.
We must also ask:
What new capabilities can this technology create?
That question is especially important with artificial intelligence.
The relationship between AI, innovation and employment was also examined on CNN’s Fareed Zakaria GPS on August 16, 2026.
Zakaria framed the discussion around the question many workers are now asking:
Is AI coming for your job?
His guests included McKinsey senior partner Asutosh Padhi and former U.S. Commerce Secretary Gina Raimondo.
Their discussion provides an important complement to the argument that innovation drives progress.
Innovation may create enormous long-term economic benefits, but societies still have to manage what happens to people during the transition.
That distinction is critical.
AI May Be A General-Purpose Technology
During the CNN discussion, Padhi placed AI alongside earlier general-purpose technologies such as electricity, computers, the internet and mobile technology.
That comparison deserves attention.
Electricity wasn’t simply a better version of an existing factory machine.
It eventually changed how factories themselves could be organized.
Computers didn’t merely replace calculators.
They transformed design, engineering, inventory management, communications, finance and production control.
The internet didn’t simply replace letters.
It transformed commerce, supply chains, purchasing, communications and the global movement of information.
AI could produce a similarly broad transformation.
And if AI truly is a general-purpose technology, its ultimate effects may extend far beyond today’s chatbots and generative AI applications.
We may still be seeing only the beginning.
What Does AI Actually Mean for Manufacturing?
Manufacturing provides an excellent environment for examining the difference between AI as a tool and AI as a replacer.
Consider a manufacturing engineer investigating a recurring production problem.
Traditionally, the engineer might have to gather production records, analyze spreadsheets, review quality reports, examine downtime information, talk with operators and maintenance personnel, compare drawings and specifications and then develop possible explanations.
AI could potentially accelerate parts of that process.
It could help summarize maintenance histories.
It could identify patterns in production data.
It could assist with documentation.
It could highlight unusual relationships between variables.
It could help engineers explore potential causes of defects or equipment failures.
AI may increasingly contribute to predictive maintenance, quality inspection, scheduling, process optimization, engineering analysis and product development.
But after the AI produces an answer, someone still needs to ask:
Does this conclusion make physical sense?
Is the underlying data reliable?
Can this recommendation actually be implemented?
Is it safe?
What happens to the rest of the production process if we make this change?
That is where engineering knowledge becomes essential.
The Factory Floor is Different From a Computer Screen
One of manufacturing’s most important characteristics is that it exists in the physical world.
Machines vibrate.
Cutting tools wear.
Fixtures move.
Material properties vary.
Sensors fail.
Suppliers change.
Operators develop practical knowledge that may never appear in a database.
Production schedules conflict with maintenance requirements.
A process that appears perfect in a computer simulation may encounter unexpected problems when implemented on an actual production line.
Artificial intelligence can analyze information describing this world.
Engineers, machinists, technicians, operators and maintenance personnel actually experience it.
That practical knowledge has tremendous value.
The future manufacturing professional may therefore need to combine two types of intelligence:
Artificial intelligence that can rapidly analyze information.
and
Human intelligence that understands context, consequences and physical reality.
The combination could be much more powerful than either one operating alone.
AI Could Become a Productivity Multiplier.
This is where the conversation about AI often becomes too focused on replacement.
Suppose AI helps an engineer complete an analysis in one hour that previously required five hours.
It is tempting to describe those four saved hours simply as labor eliminated.
But that isn’t the only possible outcome.
The engineer could use those hours to solve another production problem.
Develop a better process.
Investigate a quality issue.
Help introduce a new product.
Work with suppliers.
Experiment with new technology.
Train another employee.
Or investigate an idea that previously remained unexplored because there wasn’t enough time.
AI then becomes more than an automation system.
It becomes an innovation multiplier.
One engineer equipped with powerful analytical tools may eventually accomplish considerably more than the same engineer could previously.
That could have profound consequences for productivity.
But AI Will Replace Some Tasks
We should not pretend that technological progress produces only winners.
AI will automate work.
Some existing tasks will disappear.
Some jobs will contain fewer people.
Some occupations may eventually disappear entirely.
Other occupations will emerge.
And many existing jobs will become combinations of human and machine work.
The distinction between tasks and jobs is especially important.
A job consists of many tasks.
If AI automates 20% of those tasks, it doesn’t necessarily mean the entire job disappears.
Instead, the nature of the job may change.
Manufacturing has experienced this repeatedly.
CNC machines automated many manual machining operations, but manufacturing still requires people who understand tooling, materials, programming, process planning, quality and production.
CAD automated enormous amounts of drafting work, but it did not eliminate engineering.
Industrial robots automated repetitive operations, but factories still require technicians, programmers, engineers, maintenance specialists and operators.
AI may follow a similar pattern—although potentially at much greater speed.
The Transition Cannot Be Ignored
Former U.S. Commerce Secretary Gina Raimondo emphasized another side of the issue during the Fareed Zakaria GPS discussion: society needs to remain competitive in AI while avoiding prolonged and destabilizing unemployment.
This is where the debate becomes more complicated than simply being “for” or “against” artificial intelligence.
Stopping innovation is unlikely to be a successful economic strategy.
But neither is introducing technology without considering what happens to the people affected by it.
The goal should not be:
Protect every existing task forever.
Nor should it be:
Automate everything simply because we can.
A more productive objective is:
Increase productivity while helping people move with the technology.
That means training matters.
Education matters.
Communication matters.
And management decisions matter.
Two Factories, Two AI Strategies
Imagine two manufacturing companies introducing essentially the same AI technology.
Factory A: Replacement Strategy
The first company asks:
“How many employees can this eliminate?”
It deploys AI primarily to reduce headcount.
Employees see the technology as a threat.
Experienced workers may become reluctant to share knowledge that could be used to automate their own work.
Trust deteriorates.
The company may achieve short-term cost savings, but it risks losing valuable institutional knowledge.
Factory B: Augmentation Strategy
The second company asks:
“How can this make our employees more capable?”
It uses AI to automate repetitive analysis, improve maintenance decisions, detect quality problems earlier and give engineers and operators better information.
At the same time, employees receive training on how to use the new systems.
Workers contribute their practical knowledge.
Engineers verify AI recommendations against physical reality.
The company gradually develops something more valuable than automation:
a workforce whose capabilities increase alongside the technology.
That can create a powerful cycle:
Better technology → more capable workers → higher productivity → more innovation → stronger competitiveness.
Both factories use AI.
But they are pursuing fundamentally different strategies.
The Skills Manufacturing Workers Will Need.
The rise of AI does not mean every manufacturing worker needs to become an artificial-intelligence programmer.
But workers will increasingly need to understand how to work with intelligent systems.
Engineers may need to become better at asking questions and validating AI-generated analysis.
Technicians may interact with AI-assisted diagnostic systems.
Quality professionals may work with increasingly sophisticated machine-vision tools.
Managers may use AI for planning and decision support.
Operators may interact with systems that continuously adjust or recommend production parameters.
Across these occupations, one skill becomes particularly important:
knowing when not to trust the machine.
AI can produce an answer that sounds convincing while being wrong.
In manufacturing, a convincing but incorrect answer can have physical consequences.
Products can fail.
Machines can be damaged.
People can be injured.
Customers can receive defective components.
Human judgment therefore does not become irrelevant simply because AI becomes more capable.
It may become more important.
Human Skills Could Become More Valuable.
As machines become better at processing information, distinctly human capabilities may become increasingly valuable.
Critical thinking.
Creativity.
Communication.
Leadership.
Practical experience.
Systems thinking.
Curiosity.
Judgment.
The ability to recognize when something simply doesn’t look right.
These abilities are difficult to capture entirely in an algorithm.
A veteran machinist who hears an unfamiliar sound from a machine may recognize a problem before a monitoring system does.
An experienced engineer may immediately question an AI recommendation because it violates a basic physical principle.
A production supervisor may understand that a mathematically optimal schedule is impossible because of a practical constraint the software doesn’t know about.
These examples illustrate why the future may not belong exclusively to AI experts.
It may belong to domain experts who know how to use AI.
The Biggest Risk May Be Refusing to Adapt.
There is another side to the employment question.
Companies that refuse to adopt useful technologies may eventually become less competitive.
And an uncompetitive company does not protect jobs indefinitely.
If competitors can produce higher-quality products faster and at lower cost by intelligently using AI, companies that reject the technology may ultimately face a much larger employment problem.
The challenge therefore isn’t to avoid AI.
It is to adopt it intelligently.
Manufacturers should start with specific business problems rather than implementing AI simply because it is fashionable.
Ask:
Can AI reduce unplanned downtime?
Can it detect quality problems earlier?
Can it shorten engineering analysis?
Can it improve quoting?
Can it accelerate product development?
Can it reduce repetitive administrative work?
Can it help preserve knowledge from experienced employees?
Can it make newer workers productive more quickly?
If the answer is yes—and the benefit can be demonstrated—then AI has a legitimate business purpose.
Innovation Without People Is an Incomplete Strategy
Paul Johnson’s argument about innovation and the discussion on Fareed Zakaria GPS ultimately point toward the same tension.
Human progress depends heavily on our ability to innovate.
But innovation occurs within societies populated by real people whose livelihoods can be disrupted by technological change.
Both realities deserve attention.
The answer cannot simply be to stop technological development.
History suggests that societies and companies that stop innovating eventually fall behind.
But the answer should not be to treat workers as obsolete components waiting to be removed from an increasingly automated system.
Technology should expand human possibilities.
That requires companies to invest not only in artificial intelligence but also in the people expected to work alongside it.
AI: Tool or Replacer?
So we return to the original question.
Is artificial intelligence a tool or a replacer?
The answer is:
It can be both.
AI will replace certain tasks.
It may replace some jobs.
It will transform many more.
But replacement alone does not capture the scale of the opportunity.
AI can also give an engineer greater analytical capability.
It can give a technician better diagnostic information.
It can help an operator detect a problem sooner.
It can help a small manufacturer access capabilities previously available only to much larger companies.
It can accelerate research.
It can increase productivity.
And it can potentially free people from repetitive work so they can concentrate on problems requiring judgment, creativity and experience.
The Real Competition: Humans With AI vs. Humans Without It.
Perhaps the most important future competition will not be:
Humans versus AI.
It may be:
Humans using AI effectively versus humans who are not.
The same may apply to businesses.
The manufacturers that succeed may not necessarily be those with the most AI.
They may be the organizations that figure out where artificial intelligence genuinely improves human capability and where human judgment must remain in control.
That distinction matters.
AI should not determine manufacturing strategy.
Manufacturing strategy should determine how AI is used.
Innovation is essential to progress, but technology is ultimately a means rather than an objective.
The objective is to build better products, solve harder problems, improve productivity, create competitive businesses and expand what people are capable of accomplishing.
If artificial intelligence helps us do those things, then perhaps the most productive way to think about AI is not as the machine waiting to take our place.
It is as the next powerful tool humans must learn how to use.
