Machines at the Gates of the Workshop: Engineering, AI, and the Narcissistic Injury of Losing Control
There is a peculiar battle unfolding inside the world of open-source software, hardware communities, and independent technology companies. On the surface, it appears to be a debate about artificial intelligence. It is framed as a technical argument: Is AI-generated code good or bad? Should AI-assisted contributions be allowed? Can a machine produce something worthy of being merged into a serious project?
But beneath the technical debate lies something far older and far more human.
This is a conflict about identity.
It is about what happens when a group of people who have spent years, sometimes decades, building a rare and difficult skill suddenly encounter a technology that appears to collapse the distance between expert and novice.
The engineer looks at the AI-assisted contributor and thinks: You did not spend ten years learning what I learned. You did not suffer through the endless debugging, the obscure documentation, the nights spent chasing impossible bugs. You did not earn this.
And from the perspective of the user, something entirely different is happening.
The user is not saying: “I am now equal to you.”
The user is saying: “For the first time, I have a tool that allows me to communicate with you more effectively.”
This distinction is where much of the emotional conflict begins.
The Return of the Luddite, But in a Server Room
The modern anti-AI movement often resembles an updated form of Luddism—not the historical misunderstanding of workers simply hating machines, but the deeper psychological phenomenon of a community whose social position is threatened by technological change.
The original Luddites were not opposed to technology itself. They opposed a transformation of economic power that threatened their livelihoods and social status. The machine was not frightening because it existed; it was frightening because it changed who held power.
The same dynamic appears today.
A developer community built around expertise, craftsmanship, and technical gatekeeping suddenly encounters large language models that allow outsiders to participate. Someone who previously could not understand a complex codebase can now ask questions, generate experiments, create patches, and submit ideas.
The machine does not necessarily replace the expert.
But it does challenge the expert’s monopoly on access.
And that is a much deeper psychological wound.
The Locus of Grandiosity
Psychologists who study narcissistic structures often discuss the concept of a locus of grandiosity: the external arena where a person derives their sense of uniqueness and superiority.
For some people, that arena is wealth.
For others, physical appearance.
For others, social status.
For highly technical individuals, especially those who have built their identity around intellectual achievement, the locus of grandiosity is often competence itself.
The person does not think:
“I am valuable because I have money.”
They think:
“I am valuable because I understand things other people cannot.”
This is not inherently pathological. In fact, difficult fields require confidence, obsession, and commitment. Many great engineers, scientists, and artists possess extraordinary belief in their own abilities because that confidence allows them to attempt things most people would never attempt.
The problem arises when competence stops being something you possess and becomes something you are.
Then any technology that lowers the barrier to entry feels like an existential attack.
As Sam Vaknin has written about narcissistic dynamics, the narcissist’s deepest need is not simply admiration, but uniqueness. The terrifying possibility is not failure—it is ordinariness.
The injury is not:
“I might be wrong.”
The injury is:
“Perhaps I am not as uniquely irreplaceable as I believed.”
This is the essence of narcissistic injury.
AI Does Not Destroy Expertise. It Destroys Exclusivity.
The uncomfortable truth is that AI does not eliminate expertise.
A person who understands software architecture, security, algorithms, operating systems, and debugging will still outperform someone randomly prompting a chatbot.
The difference is that the expert now has a multiplier.
The novice also has a multiplier.
And that is the psychologically destabilizing part.
For decades, technical communities often operated like guilds. The barrier to entry was not merely learning—it was surviving the initiation process. You had to read the manuals, break things, ask embarrassing questions, and slowly accumulate competence.
That process created identity.
The person who survived the journey earned membership.
Then along comes a machine that says:
“What if we could shorten the path?”
The immediate reaction from some experts is not always about quality control.
Sometimes it is about protecting the meaning of the journey.
Because if someone can use AI to generate a functioning application, what does that say about the years spent mastering every individual step?
The answer should be: it says nothing.
A calculator did not make mathematicians obsolete.
A compiler did not make programmers obsolete.
CAD software did not make engineers obsolete.
But every one of those technologies changed what expertise looked like.
The Engineer Versus the Public
This brings us to the practical problem facing owners of hardware and software companies.
A brilliant engineering team is not automatically a good public relations team.
This is not an insult.
Engineers are often exceptional precisely because they are unusual. They are comfortable spending thousands of hours obsessing over details that most people never notice. They enjoy abstraction. They enjoy complexity. They enjoy solving problems that frustrate everyone else.
Those traits create great engineers.
They do not always create great community managers.
When engineers are placed directly in constant contact with users, they experience something psychologically difficult:
Reality.
Not the controlled reality of a codebase where every variable can eventually be understood.
The chaotic reality of humans.
Users ask strange questions.
Users misunderstand documentation.
Users submit messy patches.
Users arrive with AI-generated suggestions that are partially brilliant and partially nonsense.
For an engineer whose identity is built around being the person who understands, this can feel like an assault.
It is like forcing a person to stand permanently in front of a mirror that reflects every imperfection.
The result can be defensiveness.
Dismissiveness.
Even hostility.
The outside world interprets this as arrogance.
Sometimes it is.
But often it is a defensive reaction from someone whose professional identity is being challenged.
The Ticket Tracker Problem
This is why companies need separation between engineering and public-facing issue management.
A public issue tracker is not merely a technical database.
It is a social environment.
The user community needs someone whose job is to translate:
“This user is confused”
into:
“This product area has a documentation problem.”
And:
“This AI-generated patch is terrible”
into:
“This contribution needs technical review.”
Without that layer, engineers are forced into a role they were never designed to perform.
The danger is that asking engineers to give up direct control of tickets may itself trigger resistance.
They may interpret it as punishment.
They may hear:
“You are being removed because you cannot handle users.”
When the intended message should be:
“You are too valuable to spend your time fighting social battles.”
The goal is not to take power away.
The goal is to protect engineering focus.
The Hidden Message of Gatekeeping
The most damaging form of anti-AI resistance communicates something unintended:
“You are not qualified to participate.”
The engineer may believe they are protecting quality.
The user may hear:
“This is our private club. You are not smart enough to enter.”
This is especially dangerous because open-source communities were historically built on the opposite idea: that people from different backgrounds could contribute.
The irony is that AI may actually expand the original open-source dream.
A person who cannot memorize every programming language, understand every framework, or spend ten years becoming a specialist can now participate more meaningfully.
That does not mean their contributions should automatically be accepted.
It means they can finally enter the conversation.
The Reasonable Position
The mature position is not blind enthusiasm.
AI-generated content absolutely creates problems. The internet is already flooded with low-quality automated material. The concern that AI can produce garbage at industrial scale is legitimate.
Even the strongest AI advocates acknowledge that overuse creates problems; the comparison to technologies like antibiotics is useful because a powerful tool can become harmful when indiscriminately applied.
The answer is not prohibition.
The answer is judgment.
The strongest argument for AI is not that machines are replacing humans.
It is that humans now have access to better tools.
A competent engineer with AI assistance is still a competent engineer.
A competent writer with AI assistance is still responsible for the words.
A competent company owner with AI assistance still needs vision.
The mistake is confusing the tool with the person using it.
The Final Irony
The greatest threat AI poses to many experts is not that it will make them obsolete.
It is that it will make them less mysterious.
It will expose which parts of their work were genuinely creative and which parts were accumulated procedure.
It will separate expertise from performance.
And that is uncomfortable.
Because every profession has rituals that maintain status.
Every profession has things that seem impossible until someone builds a tool that makes them accessible.
The challenge for engineers, creators, and technical communities is not to defend the gates of Olympus forever.
The challenge is to recognize that the gods were never powerful because the gates were closed.
They were powerful because they built something worth entering.
The future belongs not to those who ban the machines at the gates, nor to those who blindly worship them.
It belongs to those who understand what the machine is actually good for—and what still requires a human being.