Educ AI tion

07 May 2024

I. Introduction

It’s clear that the introduction of AI chatbots like ChatGPT, Google Bard, etc. has completely transformed the educational experience. It is now possible to have an entire essay written for you in less than a minute. On top of this, AI can help with math, writing code, and almost anything else you can think of. While there are obvious concerns for this powerful tool to undermind student’s education, I found AI to be very useful when breaking down small parts of a larger issue I was trying to solve. I think this is an effetive and healthy way to use AI to assist in education, rather than replace it.

When I think of the future of education and what degree of impact AI will have, I am reminded of the concerns that the introduction of the internet would make people less capable professionals, and “dumb them down”. As we have seen, the educational system has adapted and embraced this change and I expect AI to be no different in due time.

II. Personal Experience with AI:

I have used AI in class this semester in the following areas:

Experience WODs e.g. E18
When working on the Experience WODs, I did not use AI often, if at all. We were provided thorough guides and documentation that helped me work through the assignments and learn the new material. I think AI could have been useful, but I felt comfortable completing the assignments and felt as though I would learn the material better if I avoided using it.
In-class Practice WODs
For the in-class practice WODs, I used AI to help with small tasks like centering an text element, formatting a background image, and other CSS tweaks. I found AI to be very useful in this regard, but less useful when trying to accomplish larger scale tasks like setting up a meteor template.
In-class WODs
I used AI here and there in the in-class WODs. When under a time crunch or when I felt totally stuck with the clock ticking down, I used AI so that I wouldn’t take a large hit to my grade. Similar to the practice WODs, I found AI useful for small tasks and the initial JSFiddle WODs that were primarily focused on JavaScript logic e.g. (“Why are my values coming back as undefined when running this block of code: … “).
Essays
I used AI to maintain proper grammmar and spelling. I did not use AI to assist with my actual writing since most of the essays were expected to be opinionated or personal.
Final project
When working on my final project, I used AI to help with some areas of creating a meteor web application that I did not feel confident / familiar with. An example of this was making sure I had subscribed to any collections I was using with the proper syntax (“What do I need to include in my Mongo.js file to make sure I am properly subscribing to the Projects subscription?”. Similar to the practice WODs, I used AI to help format a lot of the CSS work I did.
Learning a concept / tutorial
When learning a new concept in class, I felt like the screencasts and experiences were more than adequate for helping me understand new technical skills. I rarely, if ever, used AI for this reason.
Answering a question in class or in Discord
Many times in class discussions, we were encouraged to consult AI to gain a better conceptual understanding of whatever topic we were learning. I had never experienced this in a class, but I think it was a great way to have all of the main bulletpoints presented to us in an easy-to-understand way.
Asking or answering a smart-question
When coming up with smart-questions in this class, I would ask ChatGPT something like “How can I make this question worded to be consise and descriptive enough to get my point across” to improve the quality of the question I was asking.
Coding example e.g. “give an example of using Underscore .pluck”
When working on my final project, I was in charge of formatting lots of the CSS and styling in our application. I used AI for styling more than any other aspect of its creation. An example of a prompt I gave ChatGPT was “How do I set the background of this page to an image but have that image remain fluid so it expands/shrinks for different window sizes?”. Then I could model my code off of the example it provided.
Explaining code
I found AI to be helpful for explaining blocks of code that was unfamiliar to me. Especially when learning the underscore methods. I would see an underscore method used with multiple parameters, and ChatGPT was able to clearly outline what each argument was used for and what the method accomplished.
Writing code
When writing code, I primarily used AI to help me if I got really stuck or if I had spent too long trying to solve an issue. This was usually for small syntax issues or some logic that I had gotten wrong.
Documenting code
I did not use AI to document any of my code. I’m sure it could have helped me with this, but I felt like I needed to understand what my code was doing, so I made a point of writing any documentation myself.
Quality assurance
I rarely used AI to ensure quality in my code. We had other tools at our disposal such as ESLint, testcafe, and manual testing which I used to make sure my code worked properly.
Other uses in ICS 314 not listed above
I cannot think of any times that I used AI to assit me with this class other than the examples provided above.

III. Impact on Learning and Understanding:

Incorporating AI into my learning of new topics in this class was overall very helpful. I know that it saved me a lot of trouble when trying to understand new areas of software engineering. It helped me build my skills faster and overall made me a more competent software engineer. I think it is extremely important to analyze how much I utilized AI and how much I was actually learning. I think relying on AI too much would be a recipe for disaster when trying to actually join the workforce in this industry. So it’s important to manage it, but I do believe AI can be a very effective tool to enhance one’s learning.

IV. Practical Applications:

The use of AI has practical applications that extend to almost any industry in the modern era. AI can be used for improved medical analysis and interpretation, or analyzing market fluctuations for financial trading. This highlights not only the strength of some of these AI bots, but also their flexibility. Almost any business can potentially benefit from the insight of these massive data treasure troves. Within the realm of software engineering, I think AI definitely has its limitations. It’s great at giving accurate responses for small issues, but less so for larger scale problems. This could be an abstract concept, creative decisions, or even undestanding the context of an entire application’s project structure. So while AI is great for certain apsects of software engineering, there are some limitations.

V. Challenges and Opportunities:

As mentioned in the Impact on Learning section, an overreliance on AI can leave a student unprepared to face real-world problems they may face in their career. A lack of general knowledge can leave you passed over for an opportunity, even if you’re great at writing prompts to ChatGPT. Additionally, sometimes ChatGPT can write some really poor code. Sometimes it will overcomplicate ways of accomplishing a goal and create unnecessary confusion to a reader. An opportunity for further integration of AI in software engineering education might include personalized learning assistance or tutoring. If AI could read code that you write and recognize common mistakes you make, it could create custom learning modules that would help the programmer learn how to avoid these mistakes they find themselves making frequently.

VI. Comparative Analysis:

When considering of traditional learning methods vs AI-assisted learning, I tend to think of the parallel conversation of book learning vs school in the age of the internet. When the internet was still young, there was concern that having immediate access to an entire ocean of information would make people lazy, undereducated, and too reliant on external help. I think AI is just an extension of this issue. While it is an incredibly powerful tool and has the potential for abuse (cheating), the future of education needs to embrace the change and find a way to effectively use it. It’s crucial to examine the important aspects of learning: engagement, retention, and skill development. Obviously if a student can immediately have a complex essay written for them in no time at all, there is a concern of a student simply submitting the assignment that was done for them and moving on. There is clearly little to no improvement in any of these three areas. However, if AI was used to help the student formulate an outline or provide a starting place to tackle a large problem, the student could avoid any discouragement and use AI effectively to be more successful.

VII. Future Considerations:

It’s very difficult to discuss the future of education with regards to AI. This is a very new topic and while you can draw similarities to the introduction of the internet, this is an entirely different beast. AI is able to instantly write a 3000-word essay without a stutter in its step. Students have not had access to anything this powerful in human history. I don’t really know how AI would be effectively managed in terms of education whether its regulation, monitoring, restricted use, etc., but I do think it is important to make sure students don’t abuse AI at the cost of their own education.

VIII. Conclusion:

In conclusion, I believe AI chatbots like ChatGPT, Google Bard, etc., are extremely effective tools in furthering one’s education but with great power comes great responsibility. On one hand they can be a great asset within education and provide insights that may have not been considered otherwise. On the other hand, there is ample opportunity for students to skirt by in their education and with minimal effort, get through their classes, which in turn can serverly undermind their academic growth. Personally, I feel proud that I have used AI effectively to further my education without leaving me unprepared for the future.

AI was used to help with outlining and spellchecking.