Let’s cut the noise. The current freak-out over Artificial Intelligence, which promises a jobless, existence-threatening future, is fundamentally ahistorical. It misunderstands both the technology and core economic principles. AI is not some tragic, fundamental technology poised to undermine humanity; it is simply the latest, massive leap in efficiency.
Technology Has Always Replaced Labor
The core fear is that AI will take all the jobs. It’s a legitimate concern for any individual facing displacement, but let’s look at the historical precedent.
In the grand sweep of human progress, technology replacing human labor is literally the entire story. Every major innovation, from the stone tools to the plow, the assembly line, and the internet, has done exactly this. And what has happened each time? The world got richer as a result.
Take the cotton gin as a perfect example. It single-handedly replaced countless hours of human labor for seeding cotton. At the time of its invention, there were genuine fears of permanent, tragic unemployment. Despite those fears, the cotton gin made cotton wildly cheaper and more accessible. Critically, it led to the massive expansion of the textile industry, creating far more new jobs in manufacturing, distribution, and design than it ever replaced.
In every case, the positive effect of making a key commodity or service accessible to the population at large has always outweighed the initial negative effect of job displacement. AI is doing precisely this to information work. The belief that “this time is different” ignores the economic reality that efficiency has always bred new, unanticipated forms of wealth and labor.
Outsourcing Boredom, Not Brilliance
Look at what current AI is actually doing: processing data. That’s the magic trick.
Large Language Models (LLMs), are statistical engines. They crunch vast amounts of text and spit out a statistically plausible next word. They are incredible tools for pattern recognition and synthesis. The speed at which modern data centers process this information is a massive technological leap, but to call it a bad thing is absurd.
We are finally outsourcing the most tedious, repetitive, and machine-like tasks to actual machines. Imagine if humans still had to manually sort, categorize, and cross-reference a billion documents just to compute the weights in the artificial neural nets that comprise a language model. They would be incredibly inefficient, prone to error, and maybe even miserable if they have other aspirations in life.
It has always been a good thing that this kind of heavy, mechanical labor is done by machines instead of humans. We are simply realizing a colossal efficiency dividend on data processing. Complaining that these machines are too good at pattern matching is like complaining that a calculator can add faster than you can. It’s what they’re built for. The idea that advanced data processing is somehow a tragic path for humanity is a confusion of progress with peril.
The Real Fear is Local, Not Global
I hear the debate daily. In company break rooms, you overhear fears that “AI is going to be bad for society.” But if you listen closer, the real fear is much more personal and much less apocalyptic.
Take software engineering as an example. The fear engineers express is not about the entire society, but about losing a specific economic advantage they currently enjoy.
For the last decade, software engineering has been like a gold rush. Demand for talent exploded faster than the supply of people who could write reliable code. This scarcity drove up salaries and gave engineers immense leverage. I am often amazed at the wages I am paid to write a piece of code, often rather simple code, especially because I have had to do much harder tasks in other fields such as music and education for a small fraction of the pay.
Now, AI is fundamentally changing the talent supply, helping a small team of engineers produce the work equivalent of a large team from just a few years ago. The unspoken, localized fear is: “Will my work be worth as much when the barrier to entry for coding gets dramatically lower?”
This confusion largely results from regarding the economy as a fixed pie where knowledge workers must compete over a set amount of rewards. In reality, the economic pie is an expanding one. Just as the cotton gin didn’t kill the textile industry but grew the industry by making clothes affordable for everyone, AI will expand, not shrink, the software industry and the number of jobs in it.
It will make custom software affordable and accessible to millions more businesses and people who couldn’t use it before. This massive expansion of the addressable market will likely grow the overall demand for engineers, though their daily routine will certainly change. The software jobs of the future will involve less line-by-line coding and more designing, architecting, and managing complex systems utilizing AI.
The resistance, therefore, often isn’t about saving humanity; it’s about defending the temporary, highly profitable scarcity that defined the last ten years of the industry. Similar dynamics apply to other knowledge work.
Human Creativity is Irreplaceable
As machines take over the tasks of pattern recognition and data synthesis, the value of uniquely human skills will rise. This shift doesn’t make us redundant; it liberates us to focus on what only humans can do.
I have a serious problem with the inclusion of the word “art” in “AI art.” While programs can generate images, music, and movies, these systems imitate, sample, and remix content based on billions of existing examples. They do not possess the original intent, consciousness, emotion, and lived experience required for true art. Art is a meaningful expression of the human condition. I do not consider a statistical compilation of existing styles by computers to be art.
The future society in which more and more mechanical work is outsourced to machines will place its highest value on skills that resist automation: originality, synthesis, and non-linear creativity. To thrive in this environment, you must transcend the purely transactional, data-processing mindset. AI can certainly read a vast amount of text and summarize the information, but ironically, it is more important than ever that you, the human, are also well-read. Now is a great time to start delving into wide-ranging fields such as history, mathematics, philosophy, literature, music, and more to cultivate a robust internal framework of meaning.
Even if we get to the point where AI spits out the proper course of action for the entirety of humanity, it is ultimately up to us to exercise judgment and decide if we are to follow its advice. By committing to deep, multifaceted learning, we ensure that we are the essential sources of original value in a world that increasingly craves human wisdom.
