AI Takeover
12 Dec 2025
I. Introduction
During my college years, the fast-paced environment and challenging coursework made AI a steady contributor to my academic success. The same can likely be said for most people nowadays, even those outside academia, where AI has integrated into most fields of productive work. Within the scope of my courses, I primarily use Chatgpt to aid in my understanding of concepts, complete menial tasks, and double-check my work. In Software Engineering, from my experience, I’ve seen AI tools like GitHub Copilot and Chatgpt easily assist with streamlining tasks such as code generation and debugging.
II. Personal Experience with AI
1. Experience WODs
At first, I sparingly used AI tools since for the most part I had the ability to work through the experience wods. However, as the exercises became more challenging, involving complex concepts and multiple applications and tools, I began to rely more heavily on Copilot and ChatGPT to keep up with the evolving learning landscape. I mostly used AI to help with debugging when I couldn’t find the solution, or I would take a peek at the video example.
2. In-class Practice WODs
Similar to the previous section about experience wods, I didn’t use AI much at the beginning of the semester. But as time went on, I realized that in order to complete the assignments within the time limit I would need to get help from somewhere. It reached a point where I wasn’t fully understanding the material, as I would often copy parts of the prompt I didn’t know how to solve and have AI generate a solution for me. Other times I would simply use AI as an assist to help me debug.
3. In-class WODs
Regarding the official in-class wods since they were weighed so heavily, I found myself under a lot of pressure. Since the difference between an F and an A could be as small as a slight error in output, combined with the fact that we were timed led me to use AI as a crutch. After failing my first two words, I decided that I would sacrifice my understanding of the assignment in order to protect my overall grade. Like the practice wods I would often copy parts of the prompt into AI and use its debugging abilities to fix my code.
4. Essays
For the essays on this course, I would use AI to check my grammar and the flow of my structure every so often.
5. Final project
With the final project, I heavily relied on Chatgpt and Copilot while facing problems I had never worked with before. I used Copilot to generate certain parts of my issues to streamline my workflow, allowing me to focus on other, more time-intensive tasks. I also used it to help out with most of my debugging when I would run into page crashes or simple output issues. AI also played a role in helping me brainstorm approaches whenever I was stuck and unsure of what my next step should be.
6. Learning a concept / tutorial
I rarely used AI to help me understand a concept during the learning phase, since I didn’t really allocate time to actually go through the materials posted on the course website. I prioritized other tasks instead, as I found that knowing the materials didn’t significantly impact my performance in the class.
7. Answering a question in class or in Discord
For answering questions in class or discord, I didn’t use AI tools at all. But this is because I never found myself in a situation where I answered a question in either place.
8. Asking or answering a smart question
When it came to asking or answering smart questions, I rarely did either, so AI was never used. However, when I did have a question to ask, I didn’t rely on AI, since I already had a clear idea of what I wanted to ask and what I hoped to gain from it.
9. Coding example
I never used AI to find a coding example because I was usually able to easily find one through an online recourse.
10. Explaining code
When I needed to look through code to understand what it was doing, I would usually start by looking through it myself. In the case that I am unable to read it, I would just ask AI to explain it to me.
11. Writing code
There were times when I had Copilot generate code for me. In these cases, I would typically test the written code, iterating on it and sending more context until I arrived at the solution I needed.
12. Documenting code
I never used AI to document my code, as I typically do not engage in code documentation, even though it offers significant benefits.
13. Quality assurance
For most quality assurance issues I would usually take a look at it at first and assess if I could figure it out. If I didn’t know what to do I would use the quick fix option that Copilot offered or I would paste in whatever error it was popping up into the chat.
14. Other uses in ICS 314 not listed
I didn’t find any other uses of AI within ICS314 that are not listed.
III. Impact on Learning and Understanding
I believe that using AI hindered my learning and understanding greatly throughout the entirety of this course. I have a lot more experience with a wide variety of tools now, but a shallow-depth understanding in terms of the key concepts.
IV. Practical Applications
Outside of ICS 314, AI has been used in real-world projects and team activities, like hackathons, to help with code generation, debugging, and running simulations. It’s useful for speeding up work and reducing mistakes. The thing is, it works best when you understand what it’s doing, because relying on it too much can cause problems.
V. Challenges and Opportunities
I noticed especially while working on my final project, that in some cases, when you simply keep reiterating issues without understanding the why behind it you don’t get anywhere. It’s important to know yourself what your goal is and to think on your own. Relying too much on AI to fix issues in code is a bad approach and will often not lead to success.
VI. Comparative Analysis
I believe that using AI as an enhancement to learning in this course effectively killed my motivation, engagement, knowledge retention, and practical skill development. More traditional teaching methods work better for me since it allows me to actually understand the key concepts of the course. The whole point of learning a course is to gain knowledge, not to use AI to do all the work for us. And given the choice between a grade and learning, students will always choose what benefits them in the short-term.
VII. Future Considerations
I believe AI still has a long way to go before it can be effectively integrated into software engineering education. In its current form, it often acts as a supporting tool, but it can also become a significant limitation. For students to truly benefit, AI should be used thoughtfully as a supplement to active learning rather than a substitute for it.
Conclusion
Overall, my experience with AI in the Software Engineering course was mixed. While AI tools like Copilot and ChatGPT helped streamline workflows and offered guidance when I was stuck, they sometimes reduced my motivation, engagement, and development of practical skills. For future courses, I recommend using AI as a supplementary tool rather than a primary solution. It’s important to encourage students to engage with the material first and use AI to enhance understanding second.