Ever since the dawn of AI, its role in education has been widely controversial and debated. The controversy arises in the notion that using AI in classwork that have otherwise not been designed around the use of AI in mind allows students to essentially get a free pass, thereby learning nothing. This brings us to the role of AI in Software Engineering, both in the classroom and in professional environments. Some concepts and AI etiquette I’ve used in this class were: not over-relying on AI, reading the entire generated output until I fully understood the solution, and choosing to ask vague queries rather than hyper-specific ones to make a more generalized solution. At the beginning of the semester, I stuck to using both ChatGPT as that’s what I’ve had the most experience with prior to this class. I would later switch to using both Deepseek and ChatGPT to get more answers, but would end up only using Deepseek entirely.
A. Experience WODs (workout of the day) e.g. E18
I did not use AI because this was supposed to be a learning exercise.
B. In-class Practice WODs
I did not use AI because the practice WODs were very similar to the experience WODs.
C. In-class WODs
I used AI whenever I got stuck and felt that it was justified because this was an applied learning experience.
D. Essays
I never used AI for the essays because I didn’t want AI, or anything for that matter, to be skewing my thoughts in writing. I didn’t check grammar with AI either because Google Docs has the in-line grammar error detection.
E. Final project
I used AI for troubleshooting a lot of errors that came up whenever I couldn’t find a solution on StackOverflow. I also used AI to generate name ideas for our chatbot.
F. Learning a concept / tutorial
While AI is convenient, I don’t think that it suffices to be a tool to teach you new concepts. It cannot substitute documentation, videos, and articles as of yet.
G. Answering a question in class or in Discord
I used AI whenever the professor asked us to look up what something meant or to ask questions to the AI to engage in in-class discussion.
H. Asking or answering a smart-question
I never used AI to formulate responses to #smart-question. I felt like being an intermediary between AI and a classmate was sort of pointless.
I. Coding example e.g. “give an example of using Underscore .pluck”
I never thought to use AI for generating coding examples.
J. Explaining code
I didn’t need AI to explain code for this class as there really wasn’t too much complex syntax. I have used this for other classes like Algorithms, however.
K. Writing code
I did use AI to write code whenever I needed it to resolve an error.
L. Documenting code
I didn’t use AI to document code because I didn’t need it.
M. Quality assurance e.g. “What’s wrong with this code ” or “Fix the ESLint errors in ”
I didn’t use AI for quality assurance, ESLint checks did that for me.
N. Other uses in ICS 314 not listed
In a way, I also used AI to ask about careers in Software Development. I was never really interested in getting a career in Software Development, but this class definitely made me more open to the idea of it. I asked AI about the state of the job market in the technology sector, road maps I could take to increase the likelihood of landing a job, etc…
The use of AI in a class like this (or any coding class, for that matter) was very unique as there really is never a correct answer or one highly preferred way of doing things in this class. For example, in a class like calculus, there would only be one answer to the derivative of a function, however there are still many ways to get there. In this class, there are also many different ways to get to many different answers that are all equally correct. In other words, I believe that using AI has not stunted my learning progress because of how much nuance and complexity there is in software development (or any other coding class). I firmly believe that AI has in fact boosted my growth by streamlining the learning process by cutting out the parts that are menial and tedious and simplifying the parts that are too complex for my current skill level. It sort of acts like a catch-all net that makes learning concepts faster and more convenient. I still do believe that it cannot replace the human side of teaching which consists of documentation, videos, articles, and university; rather, let it be a supplement to learning.
I think that a lot of working professionals, especially younger ones, use AI. I’ve seen a lot of videos where the idea of “just feed it to the AI” was brought up whenever someone would ask a question. I think that AI has been applied and is starting to be ingrained in the daily lives of people. I’ve used AI before on a science fair project. I remember using it to help me with writing google map functionality. I was watching a tutorial video on it, but it ended up being a deprecated solution. I used ChatGPT to update the deprecated solution and I only had to make minor tweaks to fit it to my app. I think that using AI for projects, competitions, or even your daily life is a pretty good idea. It has streamlined a lot of processes for me.
I faced a challenge in one of the in-class WODs when none of the answers ChatGPT gave me got my solution to work. I didn’t try to adapt new solutions besides trying to solve it on my own without AI, but this didn’t work either. Eventually, I ran out of time and didn’t finish it. I remember putting in clear instructions to ChatGPT for a good 40 minutes, but it still couldn’t figure it out no matter how much I’ve clarified the problem. So, after class, I thought of using another AI tool: Deepseek which immediately gave me a working solution on its very first try. After this, I learned to not put all my eggs in one basket by completely relying on a single AI.
Traditional teaching methods such as university, online courses or YouTube videos, or, self-study via research can oftentimes entail a lot of dissonance between students and instructor. This can be because of things like pacing, disagreements, etc.. In AI-enhanced learning, these are often nonexistent because you are able to pace yourself. A lot of people do get frustrated whenever an AI chatbot doesn’t give them a satisfactory answer, so there could still be some disagreement. However, AI cannot be as engaging or fun as real instructors, which would actually harm student’s learning in the long term. Like I said, AI has still not advanced to the point where it can completely replace traditional teaching methods. Although, it is very possible in the future.
AI can continue to improve learning by acting as a supplement to it. If it is able to take a user’s entire directory and edit code based on that directory, I believe that it would become even more useful and be able to provide users with better solutions custom fit to them. Something that has been bugging me with the use of AI is how persistent it is in giving me the same solution whenever I use the reasoning models. I try to edit the prompt and even explicitly ask it to not give me the same answer, and it does acknowledge this during its reasoning process, however it just outputs the same undesired responses. The model I use the most is the Deepseek reasoning model.
My engagement with AI in software engineering education balanced caution with practicality, strategically leveraging tools like ChatGPT and Deepseek to troubleshoot errors, generate creative solutions, and simplify complex tasks without over-reliance. While I avoided AI for foundational exercises (e.g., WODs, essays) to prioritize hands-on learning, it proved invaluable in resolving deprecated code or brainstorming ideas, acting as a pragmatic supplement—not a replacement—for traditional resources like documentation or human instruction. Challenges, such as ChatGPT’s unresponsive solutions during a critical WOD, taught me to diversify tools, switching to Deepseek for efficiency. Though AI streamlined tedious processes and accelerated problem-solving, its limitations in adaptability (e.g., repetitive outputs) and inability to replicate human engagement underscore its role as a supportive aid rather than a standalone solution. Moving forward, enhancing AI’s contextual awareness and responsiveness to feedback could solidify its utility, provided it continues complementing, not overshadowing, the creativity and critical thinking inherent to education. By the way, this summary was written with an AI tool.