No Dummy: AI Isn't Replacing Developer Jobs - Unpacking the False Narrative
The tech industry has been flooded with bold claims about AI automating software engineering jobs. These narratives, often propagated by AI company CEOs and marketing departments, lack critical examination and evidence. This post unpacks the real factors behind job losses in the tech industry and examines why the "AI will replace coders" storyline is fundamentally flawed both technically and logically.
The Critical Thinking Crisis
False Narratives and Fear-Based Marketing
The core issue isn't about AI capabilities, but rather our collective failure to apply critical thinking to industry narratives. As software engineers, we should be the first to question sweeping technological claims, especially when they come from those with clear financial incentives.
"We have a real problem with critical thinking in America. And one of the places that is very evident is this false narrative that's been spread about AI automating developers jobs."
The argument that "AI is coming for your job" serves multiple purposes - driving investment in AI companies, creating fear-based productivity in tech workers, and shifting blame from other organizational factors. However, a clear-eyed analysis reveals more mundane explanations for job losses.
The Mislabeling Problem
The term "AI" itself is problematic when applied to large language models and other current technologies:
"These are narrow intelligence, narrow AI or even you could call it narrow machine learning. They do one task like autocomplete and then they chain together these simple tests and there is no concept of intelligence that's embedded inside of there."
This linguistic sleight-of-hand elevates simple task automation to something more mysterious and powerful, making it easier to sell grand narratives about job displacement.
Six Real Reasons Behind Developer Job Losses
1. Non-Productive Employees
All organizations contain people who contribute minimal or no actual value. When these positions are eliminated, it's not because AI has replaced them but because:
"If you fire a person that does no work, there will be no impact. And that's the bottom line is all organizations contain these kinds of people. You see it in academics, you see it in management, you see it in the technical industry."
I reference David Graeber's book "Bullshit Jobs," which categorizes roles like taskmasters, box tickers, and goons - positions that exist without providing meaningful value. Eliminating these positions isn't automating real work; it's removing unnecessary overhead.
2. Low-Skilled or Negative-Value Developers
Some developers actually reduce organizational value through their work:
"There are developers that have very low or no skills, even negative skills... If you fired someone who wrote really buggy code and replaced their buggy code with maybe another developer who used, let's say, this Fancer Auto Completion through agents to be a little bit more productive, is AI replacing that person's job? No, that person had negative value to an organization."
These developers create more problems than they solve, requiring additional resources to fix their mistakes. Removing such positions and replacing them with better tooling or more skilled developers isn't AI automation - it's simply improving efficiency.
3. Traditional Automation with Basic Programming
Software engineers have been automating repetitive tasks for decades without anything resembling AI:
"I have been automating people's jobs my entire life. I literally automate people's jobs. That's what I've been doing with basic scripts. A bash script is not AI. It has no form of intelligence. It's a for loop with some conditions in it."
I'll share a personal example from my time at Disney Feature Animation around 2003, where I replaced manual weekend maintenance tasks with simple bash scripts. This type of automation has been happening throughout the history of software development - long before "AI" became a marketing buzzword.
When I was working at Disney Feature Animation, one of my tasks was to come in on Saturdays for routine maintenance - putting in CDs, defragmenting hard drives, and other mundane tasks. I realized this was just a bash script waiting to happen. After writing some code and testing it, I created an automated system where machines could reboot, load from a network volume, and automatically run maintenance scripts. No AI involved - just basic programming that's been possible for decades.
4. Outsourcing to Lower-Cost Regions
A common business practice that predates modern AI:
"When you hear someone say, 'Oh, my job was taken by AI,' it's possible your job was taken because it was outsourced to another team or to some lower cost region. This is very common in America."
Organizations may find it convenient to attribute job losses to "automation" or "AI" rather than admitting they've simply moved the same work to regions with lower labor costs like India, Mexico, or China.
5. Routine Corporate Layoffs
Regular workforce reductions are standard practice in many tech companies:
"It's very common for large companies to fire 3% of their workforce. You hear this with Apple, you hear this with Amazon, is that fear is one of the ways that American corporations, especially toxic American corporations, motivate people."
These cyclical reductions are management practices rather than technological evolutions. Attributing them to AI provides a convenient narrative that makes the cuts seem inevitable rather than deliberate business decisions.
6. Sales and Marketing Deception
Perhaps the most direct explanation for the narrative:
"The ultimate reason we're hearing that AI is automating software engineers jobs is that a sales person somewhere, either the CEO of a company like Anthropic, who is a liar who said that software engineering jobs will be automated with agents. It's just a straight up lie."
Those with financial interests in AI adoption - from GPU manufacturers to AI platform companies - benefit from narratives about AI's transformative powers. I directly challenge industry leaders:
"Likewise, we hear open AI. This is a little bit easier because Sam Altman of Open AI is a known liar. Lies about everything. Exaggerates about everything. They're just trying to sell people this false narrative to get investment and to get users."
I've seen this recently in an article from the Atlantic, where a chief marketing officer claimed they were "automating everybody's job." When I checked his LinkedIn, I discovered he'd never worked as a software engineer - only in marketing roles. How could someone with no software engineering experience possibly know if software jobs could be automated?
Understanding What LLMs Actually Do
Narrow Tasks vs. General Intelligence
Large Language Models (LLMs) excel at specific, narrow tasks but lack true intelligence:
"Of course, linear regression is useful. K-means clustering is useful. Autocompletion is useful. Transcribing audio to video are useful, but these are narrow components. Each of these narrow components has zero intelligence."
The fundamental confusion comes from mistaking task-specific capabilities for general intelligence:
"When someone says you automated a job with a large language model, what are you talking about? It doesn't make sense."
Jobs are complex combinations of tasks requiring judgment, creativity, domain knowledge, and adaptability - attributes current systems fundamentally lack.
The Dangers of Overreliance
Rather than replacing skilled engineers, these tools may actually create new risks when used by those lacking fundamental skills:
"Much more likely the story of AI is that it is a very powerful tool that is dumb as a bag of rocks and left into the hands of the inexperienced and the naive and the fools could create catastrophic software that we don't yet know how bad the effects will be."
I've seen this pattern before:
"When people who don't have basic algorithmic understanding or basic DevOps skills get introduced to high level languages like Python. I've seen them create hundreds of millions of dollars in damage to organizations."
Key Benefits of Critical Assessment
-
Accurate Problem Diagnosis: Understanding the real reasons behind job transitions allows organizations to address actual issues rather than chasing technological fantasies.
-
Better Resource Allocation: Companies can invest in meaningful improvements rather than overspending on AI solutions that may not address their actual challenges.
-
Empowered Engineers: Software developers who understand the real capabilities and limitations of current tools can use them effectively while maintaining their career security.
Conclusion
The "AI will replace coders" narrative fails basic scrutiny when examined through the lens of critical thinking. Current AI technologies offer powerful but narrow capabilities that can enhance productivity when used appropriately - but they're far from the job-replacing revolution that marketing hype suggests.
As I forcefully conclude:
"That's really the problem with the narrative that AI will replace coding jobs is there's absolutely zero evidence. In fact, there's the opposite evidence."
The most valuable approach for both individual engineers and organizations is to develop an accurate understanding of what these tools can and cannot do, use them appropriately to enhance productivity, and maintain the critical thinking skills to separate marketing hype from technological reality.
# The reality of automation isn't mysterious AI, but simple scripts like:
for machine in $(cat machines.txt); do
ssh $machine "sudo reboot && hold_in_key" &
done
wait
echo "Machines rebooted with network boot - no weekend work required!"