Are We Becoming ‘Intent Designers’ Instead of Coders? The Shifting Role of Software Engineers
Hey fellow devs!
With the rise of sophisticated AI coding agents that can generate entire microservices in seconds, a common question is echoing through the tech industry: What is the role of a software developer today?
If AI can write the syntax, generate the tests, and compile the code, are we still coders? Or are we transitioning into something different—something we call “intent designers”? Let’s explore this shift, the potential pitfalls, and why programming fundamentals are more important today than they have ever been.
The Shift to Intent Design
An intent designer is a software engineer who focuses on defining what the system should do, why it should do it, and how the architecture should be structured, while leaving the execution (typing the code) to AI agents.
This shifts the developer’s primary value away from memorizing syntax and method signatures toward: → Designing system topology and microservice boundaries. → Analyzing user workflows and system constraints. → Documenting precise technical requirements in natural language.
During our recent podcast session, Luis Rafael raised this exact question:
Luis: “How does spec-driven development change the role of the developer? Do you think, for example, we are becoming more like intent designers than coders?”
Yugo responded with a clear breakdown of the new paradigm:
Yugo: “Writing code, I think, is less about typing. You need to understand the intention and the purpose of that feature, and that’s way more powerful. We always try to make our developers stop and think: we don’t just blindly follow and start implementing stuff; you have to understand what it is solving for our users… So they have to stop and think first about what they need to do, what they are actually trying to do.”
The Trap of the Auto-Generated Loop
But being an intent designer doesn’t mean you can just write a brief prompt, step away, and let the AI do everything. One of the biggest challenges with AI agents today is their non-deterministic nature. If you don’t provide a precise specification and remain active in the loop, AI agents can easily get stuck in infinite, expensive debugging loops.
Yugo shared a great example of this from his own experiments:
Yugo: “I tried to run it and it failed because it couldn’t find the HTML class. And that’s why we still need some developer behind—a software engineer behind—to debug and say, ‘Okay, that’s what I will ask you to fix.’ We need someone in the loop to understand what’s going on, even with the help of the AI.”
AI can generate code, but it cannot understand when it has made a logical mistake that propagates through your system. It needs a human developer to audit the changes, find the bugs, and redirect the agent’s focus.
Why Fundamentals Matter More Than Ever
Because AI-generated code must be audited, coding fundamentals have become your superpower.
There is a dangerous belief that because AI can write code, beginners don’t need to learn frameworks or programming basics. The truth is the exact opposite. If you do not understand how code works, you cannot audit the AI’s output. You will be a passive observer of a black box you cannot control.
As Yugo emphasized:
Yugo: “Beginners developers should continue learning frameworks, how to code, so they can question what AI is generating. They need to learn these new tools… but the fundamentals are what allows you to test and validate the output. We need professionals who can deal with security concerns, mature system designs, and validation.”
Upskilling for the Intent Design Era
To thrive in this new landscape, you should shift your learning goals: • Master System Design: Focus on how databases scale, how caching works, and how services integrate. • Develop Auditing Skills: Practice reading and reviewing code. You will spend more time reading AI-generated PRs than writing code from scratch. • Improve Technical Writing: Your ability to guide AI depends on how clearly you can write your specifications. Precise English is your new coding language.
The role of the developer isn’t disappearing; it is maturing. We are moving from syntax-writers to technical directors.
What are your thoughts? Are you focusing more on system architecture and code auditing this year? Share your experience in the comments!
Watch the full video here for our detailed conversation.
Happy coding!
Cheers,
Pedro Cavalero
