Will AI Replace Programmers? An Honest Conversation
A personal interview on fear, opportunity, and smarter workflows
Q: Everyone keeps talking about AI replacing programmers. Should we really be worried?
A: I don’t think the “robots take every job” picture is the right way to look at it. To me, it feels a lot like when calculators first showed up. People stopped doing long division on paper, but mathematicians didn’t suddenly become obsolete. Instead, they were freed up to move to a higher level of abstraction and solve far more interesting problems.
You can also look at it through the lens of photography. Think about the shift from the physical darkroom to the digital one. The painstaking, mechanical craft of using chemicals and timing exposures was automated. Did this replace photographers? No. Their role transformed. The required skill shifted from working with enlargers and chemicals to mastering new digital tools like Photoshop and Lightroom. This new toolset allowed them to focus more directly on the art of the image: things like composition, lighting, and storytelling. Their value shifted from a purely technical process to a creative vision amplified by digital skills.
That’s exactly what’s happening here. AI is getting incredibly good at taking over the boring parts of our jobs. The boilerplate code, the project scaffolding, the basic unit tests, the simple CRUD apps. But the messy, truly human parts of software engineering? The vision for the project? Things like translating unclear business requirements, navigating the trade-offs between speed and cost, or figuring out how to keep a ten-year-old codebase from falling apart… that still requires a real person with real experience.
Q: Some folks say only “bad programmers” will be replaced. Do you agree?
A: That’s a popular take, but honestly, I think it misses the point. AI doesn’t care if you’re a “good” or “bad” programmer; it’s not judging you. It’s a machine that’s ruthlessly efficient at automating easy, repetitive tasks.
So, if your work is mostly just stitching a few APIs together or following online tutorials step-by-step, then yes, that part of your job is likely to shrink. The real distinction isn’t between good versus bad programmers, but between having shallow versus deep skills. Knowing the syntax is shallow. Knowing how to design a resilient, scalable system, and why you’re making certain architectural choices, that’s deep.
Q: How can programmers adapt and stay relevant?
A: It really comes down to shifting your focus. First, you have to lean into the parts of the job that are uniquely human: system design, navigating ambiguity, and really understanding the business context you’re building for.
And that means you need to get comfortable using these new tools. Let Gemini, Copilot, or ChatGPT handle the repetitive scaffolding. Save your precious brainpower for the hard stuff: architecture, complex debugging, and the user experience. This also means you have to get good at prompting and validating the AI’s output. You’re the expert in the driver’s seat; you need to give it clear context and always sanity-check what it gives you back.
Ultimately, it’s about expanding your view. The more you learn about DevOps, security, different frameworks, and product thinking, the more you see the bigger picture. And never, ever underestimate the value of talking to other humans. Communication, empathy, and explaining complex trade-offs to non-technical folks are superpowers AI simply doesn’t have.
Q: Can you paint a day-in-the-life picture?
A: Sure, let’s make it real. Imagine you’re building a new notification system.
Your morning is probably at a whiteboard, sketching out the architecture: the message queues, the database tables, the retry logic. That’s all you, because an AI can’t fully grasp the unwritten rules and specific constraints of your business.
By late morning, you’re ready to code. You ask an AI assistant to generate a basic message queue consumer in Python with retry logic and logging. The skeleton code appears in seconds, saving you 20 minutes of tedious typing.
In the afternoon, you do the real work. You wire that generated code into your actual system, handle the tricky edge cases the AI would never anticipate, and integrate it with your company’s unique deployment pipeline. The AI is your pair programmer, helping you debug syntax and write unit tests faster.
By the end of the day, you’re documenting the feature and talking to stakeholders. You make the strategic call to launch without push notifications until the next sprint. That’s a human trade-off, not a technical one.
In that whole scenario, the AI didn’t replace you. It just cut away the grunt work and let you focus on the meaningful, creative pieces of the job.
Q: So where does that leave us?
A: AI won’t simply “replace bad programmers.” It’s redefining what it means to be a programmer. If your contribution is limited to the grind of typing out syntax, that work will inevitably matter less over time. But when you pair your coding skills with creative problem-solving, thoughtful design, and the ability to steer these powerful AI tools, you become far more valuable than before.
Think of AI as a set of power tools. When you know what you are building, those tools make you unstoppable. Without a clear plan, they just help you make mistakes faster.
Bonus
- Follow for more educational content and stuff!
- It costs nothing to support a creator. Just hold down that 👏 button until you see 50 pop up.
- For more things that I do, visit https://bento.me/thecodemancer-davidregalado.
