AI isn’t replacing developers, it’s promoting them.
Doug Compton explains how AI is redefining the role of today’s developer, from builder to orchestrator. This isn’t automation for the sake of speed. It’s a new operating model for how software gets made.
Modern devs are becoming AI architects, quality leads, and workflow managers. The code still matters. But how it gets written? Completely different game.
Transcript
You know, typically your project manager would prepare your requirements. They would hand it off to you. And then as a developer, you would look at it and see if you had any technical questions, and then you would just start working on it. And you would, you know, work it until it was completed.
Now, whenever I get a task from my project manager, I start doing a planning session. I work with it, sort of like a collaboration as a team, have it understand what the problem is. And then we come up together with an implementation plan. And then once that’s done, then I turn the AI loose and let it write all the code. And then at that point in time, I am free to do other things.
As AI becomes more and more advanced, I find that I am more of a manager of AI. I’m also more of an architect. I need to decide what the overall structure of the code should be. Now I need to manage, you know, a couple of AIs to have them do different work. I do review all the code they write to make sure that it’s accurate and doesn’t introduce security issues.
The developer’s tools that have been evolving too, tools like Cursor and even, you know, Visual Studio with Copilot, the tools have been getting better, but also the AI in them has been getting better as well. Initially, it was more of a junior developer. It, you know, grew to where it was more of an assistant. And now it’s growing to more where it’s actually almost an equal, almost not quite, being maybe a mid-level developer where it can do a lot of things on its own.
So as AI is getting quicker, that puts a lot more work on the developer to be able to make sure everything is working okay. You might also have other AI, different agents, that will do code reviews or do security reviews that you can then monitor to make sure that the quality is good. The AI can write unit tests and integration tests. It can run those on its own to make sure that the code still works as expected.
But also it’s still up to the developer to make sure that the code that’s being written is, you know, good. So they still have to look at the code.
Working with product design has also evolved. When you start giving them the developer tools with AI, they’re able to actually do some of the front end development work. The developer probably still needs to make sure that the code that it’s writing is still good quality, do a code review on that as well. And also that gives you, a developer, a better picture of what it is they’re envisioning.
So a developer just starting out, you know, they still need to understand the technical foundations of the software stack they’re using. They still need to be the ones that direct the AI. So they need to have that knowledge, and they need to understand the limitations of AI and what it can and can’t do, and also how to prompt it to get it to do things properly.
So it’s a combination of what traditional software development was with the, knowing the technical side, but also knowing the AI tools and how to interact with it is very important as well.
Want More AI Workflow Videos?
Written by: Savannah Cherry
Savannah is our one-woman marketing department. She posts, writes, and creates all things Slingshot. While she may not be making software for you, she does have a minor in Computer Information Systems. We’d call her the opposite of a procrastinator: she can’t rest until all her work is done. She loves playing her switch and meal-prepping.
Expert: Doug Compton
Born and raised in Louisville, Doug’s interest in technology started at 11 when he began writing computer games. What began as a hobby turned into his career. With broad interests that range anywhere from snorkeling, science, WWII history and real estate, Doug uses his “down time“ to create new technologies for mobile and web applications.
Frequently Asked Questions
Traditionally, a project manager would prepare requirements, hand them off, and the developer would start working until the task was completed. Now, the process starts with a planning session where the problem is understood as a team and an implementation plan is created before AI is used to write the code.
The developer becomes more of a manager and architect. They decide the overall structure of the code, manage multiple AI tools doing different work, and review all output to make sure it is accurate and does not introduce security issues.
Tools like Cursor and Visual Studio with Copilot have continued to improve. Initially, AI behaved more like a junior developer, then an assistant, and now it is approaching the level of a mid-level developer that can handle many tasks on its own.
No. AI can write unit tests and integration tests and run them automatically, but it is still the developer’s responsibility to review the code and ensure it is good quality, works as expected, and does not create security risks.
Developers still need to understand the technical foundations of the software stack they are using. They must know how to direct AI, understand its limitations, and know how to prompt it properly. Modern development combines traditional technical knowledge with strong AI tool skills.
