Will AI Replace Programmers? Here's My Take.

The development of LLMs (Large Language Models) like ChatGPT and Claude has made it exponentially easier for large amounts of code to be generated in short amounts of time. What might've once taken a team of professional programmers hours can now be accomplished, by any average person, in less than ten minutes––and all you need is a simple prompt.

Rightfully, many are nervous about these new AI systems. If one prompt can accomplish the same and more as hired employees, then what use are human workers?Indeed, we are already beginning to see shifts in the corporate world, with workers being laid off left and right, and even less getting hired for the job in the first place. Even in universities, fewer and fewer students are majoring in sectors like computer science––one study from Yahoo finance states that while computer science used to be a student hotspot, its currently seeing the largest enrollment drop of any major in six years.

Still, I believe that there is reason to be optimistic. AI may be replacing jobs, but is also enabling new ones and making us, as a society, more productive. At the end of the day, AI is a useful tool that can perform monotonous tasks exceptionally well. But didn't this also occur in the 1900s, when machinery like tractors replaced manual crop cultivation––or when Henry Ford's assembly line replaced car craftsmen? We no longer required crop collectors, yet needed someone to drive the tractor. We no longer required humans assemblers, yet needed someone to maintain the machinery. In the same way, AI will open up new opportunities––it can't function without human help.

It's true––AI is not autonomous––and it's proven on the regular. Even in my own experience building ClassPulse, classroom software that I coded with help from Claude Code (an LLM specialized for programming), there were many moments where, without human intervention, Claude would have gone down the wrong path––putting the entire program at risk. There were even more moments where it simply couldn't detect fundamental problems in the system. It took me months to realize that I didn't have separate development and production servers, a crucial flaw, and one that Claude never pointed out. Even when Claude does execute a command, it often creates problems in other sectors––two steps forward, one step back. Without a human monitoring the AI's behavior, guiding it in the right path and detecting design failures, nothing can be accomplished.

AI simply can't stand alone right now, and it is uncertain that it ever will. "AI inbreeding," otherwise known as the Habsburg effect, is a term that describes hallucinations or low quality output from AI systems that occurs due to data collection from other AI-generated sources. Building LLMs requires a massive input of human content from media on the Internet––that's what makes AI so good at replicating human behavior. When it runs out of human content to examine, it resorts to using other AI-generated information or content, which is currently scattered across the Internet. This has a prominent degrading effect on AI output––something that humans need to pay close attention and monitor consistently.

All of this is not to say that artificial intelligence doesn't pose any threat to our society. While it will open up new opportunities, it may not do so fast enough to account for a population of unemployed, and it certainly cannot compensate for the time and energy spent on degrees that may now be rendered useless.

Additionally, LLMs don't just threaten jobs, but the actual environment we live in. Data centers being built in countless regions are draining freshwater resources, polluting the land, and putting a massive health burden on nearby citizens. The people are angry, and Mother Nature will soon be, too. The future of AI is uncertain.

But if there's anything I want you to take away from this, it's that society can adapt to this technological revolution––and so can you. Universities are already seeing shifts in coursework, with AI-centered lessons being introduced and old-fashioned coding classes being reduced. Employers are opening new sectors for AI usage and management. And you, the human, can capitalize on skills that AI will likely never be able to replicate––creativity, leadership, and passion.

Credits:

Vesovski, Victoria Vesovski. “Computer Science Was Once a “Golden Ticket” — Now It’s Seeing the Biggest Enrollment Drop of Any Major in 6 Years.” Yahoo Finance, 19 Apr. 2026, finance.yahoo.com/sectors/technology/articles/computer-science-once-golden-ticket-140500823.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAKkARUVCGoMowykPLTPNFX_KFkVNMPv_LpkoiZlclPlRMgtxu6SJq82HYPw5l8j9sWZavWZAW0vwX3akMFZGpsnpCSCKBUbM2o-0gafD-0fSsA7s84JqWMWnyWdqo7SPZnyvD3WniQSfPbjP83xvOuq2-WEeJnusH7AAoPlYVWJI.

Warrender, Emily. “What Does It Mean to “Know” Something in the Age of AI?” Open Access Government, 25 June 2025, www.openaccessgovernment.org/article/what-does-it-mean-to-know-something-in-the-age-of-ai/194647/.