It’s been an exciting journey integrating AI into my .NET Web Development course at Eastern Washington University. Building on my earlier posts about our initial experiments—“If AI Writes the Code, What Should We Teach?” and “A Powerful Hope for the Future”—I’m thrilled to share that students are still delivering outstanding results. Their original web apps—complete with front-ends, back-ends, databases, and Azure deployments via CI/CD—are innovative, polished, and a testament to how AI amplifies human creativity and engineering. A few examples include projects like Shop, Stable, and Doobles.
That said, as we dive deeper, some new challenges are surfacing, even with our solid DevOps processes in place. Here are the key areas we’re focusing on to refine our approach:
- Requirements Remain King: The toughest part of software engineering isn’t the build—it’s nailing down what to build. This iterative dance between human insight, practical needs, design artistry, and solid engineering is where the magic (and the struggle) happens. AI helps, but it can’t replace that human fusion.
- Guiding AI on Architecture and Decisions: Left unchecked, AI often opts for quick, sloppy solutions without considering long-term maintainability. This can lead to unnecessary churn, bugs, and rework. We’ve learned that providing upfront guidance on software architecture and core technical choices is essential for creating robust, future-proof applications.
- Test, Test, Test: It’s tempting to skim through testing when AI generates code so effortlessly, but we’ve hit roadblocks from complacency here. Thorough, robust testing is proving absolutely core to success. This feels like a broader unsolved puzzle in the AI world right now—and at IntelliTect, we’re actively tackling it head-on.
On a brighter note, it’s incredibly encouraging to hear from educators across the US reaching out to collaborate. Together, we’re exploring smart, creative, and humane ways to evolve teaching for an AI-powered workplace.
Designing AI systems you can rely on
If you are exploring how to use large language models in real software, IntelliTect helps you turn clear intent into reliable, production-ready systems.