Artificial Intelligence (AI) has rapidly become an essential tool in education—especially in software engineering, where fast iteration, problem-solving, and code development are crucial. In ICS 314, AI tools provided new ways to approach coding challenges, learn concepts, and accelerate development. Throughout this semester, I mainly used ChatGPT as my AI assistant, while occasionally trying GitHub Copilot and Cursor. These tools were particularly helpful for debugging, clarifying syntax, and generating code snippets.
For the Functional Programming WOD (E18), ChatGPT helped me understand how to use Underscore.js methods like _.pluck, _.reduce, and _.filter. I’d ask:
How do I use _.pluck to extract values from an array of objects?
What’s the difference between _.reduce and JavaScript’s native .reduce() method?
AI clarified syntax and use cases, but I still needed to write and debug the core logic myself.
During WODs for recreating websites (like Island Snow), I asked ChatGPT questions such as:
How do I make a navbar that collapses with a toggle button using Bootstrap 5?
How do I center an image in Bootstrap without it stretching?
This saved me time compared to searching through the full Bootstrap docs, especially under time pressure. However, I always checked results in the browser and adapted code as needed.
For the “wod-aloha-beer” project, ChatGPT helped debug HTML/CSS in real time:
How do I space Bootstrap nav links evenly across a row?
Why is my image not aligning using float-start in Bootstrap 5?
AI sped up my workflow, but sometimes gave generic answers that required manual adjustment.
When writing essays (like the Salvage Public reflection), I prompted ChatGPT to:
Help me structure a reflection essay on recreating a website using Bootstrap 5.
How can I explain my Bootstrap design choices in a student-friendly tone?
ChatGPT helped organize my scattered notes into readable sentences and conclusions, which I then personalized further.
For our group’s UH Marketplace app, I contributed to the Prisma schema and profile page. ChatGPT assisted with:
How do I define a one-to-many relationship in Prisma for users and items?
How do I fetch and display related data in Next.js using Prisma?
How do I use Tailwind CSS to align sections on a dashboard layout?
AI made it easier to get started, letting me focus on styling and feature logic.
To understand layout containers in Bootstrap, I asked:
What’s the difference between .container and .container-fluid in Bootstrap 5?
When should I use containers vs. rows and columns?
This led to better use of layout patterns in my projects.
For documenting code, I had ChatGPT generate JSDoc comments:
/**
* Returns all marketplace items for the logged-in user.
* @param {string} userId - The user ID to query.
* @returns {Promise<Array>} - List of user items.
*/
For ESLint and TypeScript issues, ChatGPT offered suggestions on missing return types, prop interfaces, and React hook dependencies, making my code cleaner and more compliant.
ChatGPT also helped brainstorm file structure and feature ideas, for example:
What’s a good folder structure for a full-stack Next.js app using Prisma and Tailwind?
It even generated placeholder item data for early UI prototyping.
AI didn’t replace the learning process—it accelerated it. I could overcome minor obstacles faster and focus on deeper concepts and architecture. However, I realized it was easy to become passive if I relied too much on AI suggestions, so I made sure to review and understand all generated code.
Outside of ICS 314, I’ve used ChatGPT in other classes (like ICS 212) and hackathons (HACC). For example, it helped me parse command-line arguments in C by explaining argc and argv clearly. This demonstrates how AI tools are becoming cross-functional helpers for CS students.
Challenges:
Opportunities:
Compared to traditional learning, AI-assisted learning is faster and sometimes more engaging—you get feedback instantly and can explore many solution paths. Traditional methods, however, often foster deeper retention and understanding. The best approach is blended: learn independently, then use AI to check, reinforce, and explore alternatives.
AI will likely become even more integrated in software engineering education, perhaps as a default “co-pilot” for students. The next step is making AI more context-aware and able to detect conceptual misunderstandings, not just syntax errors.
Using AI in ICS 314 has been a genuinely valuable experience. It helped me learn faster, write cleaner code, and find alternative solutions. But its value depends on how intentionally it’s used—AI is best seen as a partner, not a crutch. The most