Daniel Lemire

@lemire

·

5 Std.

I started programming decades ago when I was 12 years old. At first, it was just a game to me. Over time, as I acquired more skills, it became part of my identity. While others played basketball or the violin, I wrote code. It is so deeply ingrained in my identity that when I once stopped programming regularly, I became depressed. Shamefully, it took me years to realize that I need to program weekly to stay happy. Today, the best AI models can write C++ code better than most computer science professors. They are certainly better programmers than my teenage self, who learned without access to the internet. Why play the violin when anyone can find a recording of a great violinist with little effort? This is not a new problem for human beings. What happened to bowmen when the musket became standard military issue? Do you keep showing up on the battlefield with your bow? A bow has advantages over a musket—and bows continued to improve even after the musket’s invention. Yet the writing was on the wall: the future of the battlefield did not belong to the bow. I could also recount the story of portrait painters who struggled with the rise of photography. Programmers are not alone in facing a transformation of their craft. Graphic designers, lawyers, writers, and nearly everyone in the laptop class are impacted. Some do not mind, because their craft was always primarily about achieving professional success. Others are more troubled. Let us be fair. For the last 30 years, the West pursued globalist policies that continually moved industrial jobs elsewhere, leaving many young men without much-needed opportunities. We told them not to worry—their economic losses were more than offset by government checks. And this gets to the core of the issue. Getting the job done is only part of what we care about. Perhaps more importantly, we want meaningful lives, which often come through skill development and its fruitful application. Money alone is not enough—it never was. We also crave status, recognition, and a sense of empowerment. As I look at software programming today, I can accomplish in five hours what would once have taken me three days. Sometimes I can do in 15 minutes what used to take three hours. Other times—especially when tackling something difficult that plays to my strengths—I do like Luke Skywalker: I turn off the targeting computer and code without AI for the best results. At the same time, I see many people (including some of my students) trapped in an illusion of competence. They prompt an AI and generate low-quality code—not because the models are bad, but because a tool is only as good as the person wielding it. What is coming? I cannot predict the future with certainty, but the following seems possible: 1. We may end up with the equivalent of a laptop-class rust belt. People who used to write reports and PowerPoint decks for a living could become obsolete. Teachers might find students turning to AI instead, reducing them to little more than relatively useless bureaucrats. IT specialists might see their entire purpose wiped out by a new AI model. They may still earn a decent living, but their social status could crater—just as it did for factory workers in the West from 1980 to today. What goes around comes around. 2. During the globalization era, the financial industry boomed in the West, and to this day some of the most prestigious and highly paid jobs are in finance. A similar phenomenon is likely unfolding in this AI era. Many of those benefiting will have a background in programming. The people who can build in a month what once took a year will be able to automate tasks that previously weren’t worth automating. What about the sense of purpose and competence? Speaking for myself, it is not gone. If anything, I feel I can leverage my hard-earned skills even more effectively than before. More importantly: I am still having a lot of fun programming.