Chad Maeshiro

AI is Everywhere

21 Nov 2023

Introduction

Once a science fiction storyline, AI has grown to become a major industry that is expanding into a phenomenon that made its way to the general public. This expansion of AI could improve our ability to learn, but when used incorrectly, could also plague us with bad habits of getting the answers we ask for. I will explore through the usage of AI as a computer science major, the growth of AI and the advancements and limitations that it provides.

Personal Experience with AI

So far throughout the semester, I have sometimes applied or chosen not to apply AI in multiple areas of study and coursework. I have learned that the application of AI can vary from usage to usage, and I will explain in this section how AI is used differently in different forms of study and learning.

  1. Experience WODs
    • I have used A.I. for some components of my learning in my Software Engineering class this semester. In this class, we complete these assignments called ‘WODs (Workout of the Day)’ which are small fragmented coding assignments that we do in class, for homework, or as coding practice. For the homework assignments, I rarely used A.I./Chat GPT to complete them because, within the modules that the professor gave us, there were resources and videos that helped us understand how to complete those WODs. I would sometimes use Chat GPT to look up information about certain aspects of the practice WODs that I had a hard time understanding, making it much easier for me to complete them the second time around. For example, I once used ChatGPT to explain the className attribute for React Bootstrap, making it easier to work with when completing the next WOD.
  2. In-class Practice WODs
    • The In-class Practice WODs, were mainly for us to get an understanding of how the weekly WOD would be conducted. So as a practice assignment, I found rarely a need to use A.I. in these scenarios.
  3. In-class WODs
    • The In-class WODs were the weekly tests that we used in my Software Engineering to prove that we understood the concepts that we were learning throughout the week. Since these WODs tend to be more challenging than the practice and experience WODs that we’ve done in the week, I have used Chat GPT a few times during these ‘tests.’ Especially since these WODs were time-dependent on their completion, Chat GPT made it easy to look up easily forgettable information such as syntax and the occasional bugs in code. This is what I primarily used A.I. for when completing these assignments, just so that I can efficiently produce code without having to dig through documentation to get it working. For example, during a WOD that required the creation of buttons, I looked up how to stylize the button, which directed me to the correct syntax to create the button style I needed. I felt the usage of A.I. in this way didn’t impact my learning of the material, but aided my understanding.
  4. Essays
    • In the Software Engineering class, we were required to write essays to enable us to reflect on our learning. Since these didn’t require much outside information and understanding, I never used A.I. to produce these essays. However, I’ve seen other peers’ essays and how they used A.I. images within their essay pages, which I thought was very creative.
  5. Final project
    • For the class’s final project, we are designing a website that condenses information about Hawaii healthcare providers into a simple, stress-free application. For this assignment, I have used a bit of A.I. to help aid my learning when navigating through React Bootstrap. For example, the concepts of UseState were better clarified when asking ChatGPT the best use cases and how to best implement them into my website.
  6. Learning a concept/tutorial
    • I’d assume that the most common usage of Chat GPT and similar A.I. generative text is for their ability to define ideas and teach its users tutorials. I found the information that A.I. can produce intriguing, and quite amazing how it is able to take information from online and masterfully output an explanation. As mentioned above, there were many instances where I allowed Chat GPT to define or explain concepts that I had a hard time understanding. Yet to ensure that they explained the content accurately, I went to Google and compared the A.I.’s retorted explanation to answers online.
  7. Answering a question in class or in Discord/Asking a smart question
    • There has never been an instance where I felt I needed to use A.I. to answer a question in class or on Discord. Similarly, when asking for help on Discord or on Stack Overflow, there was never an issue that required the usage of AI for me to compile my question. I think this is simply due to the fact that when asking a smart question, it is best to know what you did and the issues you ran into. It would be harder for a computer to understand the issue you are running into if you yourself don’t.
  8. Coding example e.g. “Give an example of using Underscore .pluck”
    • I have actually used ChatGPT multiple times to give an example of a fragment of code. What I particularly like about doing this is that by giving an example of a certain concept I don’t understand, it enables me to understand the example it gives me, and apply that to my own code. This style gives us a more self-directed stance on learning, rather than inputting your issue into Chat GPT and the AI spewing the answer for you. In these cases, the usage of AI was very effective in supporting my learning
  9. Explaining code
    • As stated in previous sections, after understanding a concept with the aid of Chat GPT, I should have a proficient grasp of that specific subject. So when implementing it into my own program, I can write my own comments in the program myself. Using AI to explain code would only prove beneficial to someone if they either copied the code directly or failed to understand the concepts.
  10. Writing code
    • Using AI to write code has many weaknesses, which influences me to not use the code that ChatGPT produces. In order for AI to produce workable code, you have to make sure you give clear, detailed instructions. If not, you get code that won’t always work the way it intended to. For example, if you want to create a Form, you have to be sure that you explain what language you want to use, what framework to implement, the aesthetics of it, etc. Without these minor details, you don’t get the result you were hoping for. Luckily, this forces us to do the work ourselves when we fall into asking AI to do it for us.
  11. Documenting code
    • There was rarely a use case where I needed AI to document code for me.
  12. Quality assurance
    • I never used ChatGPT to fix my code for me because it feels like cheating your way out of learning. If I am desperately in need of an answer, I consult Stack Overflow for answers. For ESLint errors within my code, IntelliJ itself handles the issues really well by providing options to fix the ESLint errors directly on the code editor. This convenience prevents me from using outside technology such as ChatGPT to fix ESLint errors.

Impact on Learning and Understanding

Using A.I. within an academic setting can either help you beneficially or handicap you. What I learned while using ChatGPT is that the evolution and mass growth of the A.I. industry can make information so easily accessible, but enables us to be just as lazy. For example, when I first learned to create a navigation bar using Bootstrap, I could easily look up: “create a navigation bar using Bootstrap and Java” which will immediately showcase a great navigation bar and all the features you can add to the nav bar. Yet, all the code that Chat GPT wrote for me was easily able to be copied and pasted into my own coding assignments. This process of asking questions and retrieving answers would enable me to not learn from what information it provided for me. So doing so, I would need to go back a learn what I had skipped over myself.

Yet, AI such as ChatGPT excels at condensing information and reproducing it in an easy-to-understand format. Practical Applications Besides using AI in my Software Engineering class, I also use Chat GPT to do research for essays and other reports. Throughout this semester, I felt that the usage of AI has vastly improved my efficiency when completing these types of assignments. It is particularly useful in condensing information into understandable chunks of data. Here is an example of how I used ChatGPT to research topics for an essay in my Asian Studies class:

Q: What are some ethical considerations for the Edsa 2 Revolution?

ChatGPT:

The EDSA II Revolution in the Philippines, which took place in January 2001, involved the peaceful ousting of President Joseph Estrada and the installation of Vice President Gloria Macapagal-Arroyo as the new President. Several ethical considerations are associated with this event:

Respect for Democratic Processes: One of the significant ethical considerations revolves around respect for democratic processes. Critics argue that the ousting of a democratically elected leader might set a problematic precedent, undermining the democratic system. Ethical questions arise about the appropriate channels for addressing dissatisfaction with elected officials without resorting to extra-constitutional means.

Rule of Law: Ethical concerns arise regarding the adherence to the rule of law. The impeachment trial against President Estrada was halted, prompting questions about the integrity of legal processes. Ethical actions would involve ensuring that legal procedures are followed impartially and justly, even in cases involving high-ranking officials.

Role of Civil Society: Ethical considerations involve the role of civil society in popular uprisings. While collective action is a fundamental democratic right, ethical questions arise when assessing the influence of different interest groups, ensuring that the interests of all citizens are represented fairly.

Challenges and Opportunities

AI such as Chat GPT makes it very easy for its users to navigate and understand, acting like a simplified version of a Google search. With that, there was very little learning curve in using the website, other than having to create an account. The challenge with the usage of AI is understanding the right questions to ask, and using it for the right purposes.

Comparative Analysis

AI chat has provided a dissimilar method of learning compared to traditional teaching methods. For example, if used incorrectly, AI provides answers to questions so efficiently that the user doesn’t even have to understand the answer that they are presented with. It also doesn’t quote specific sources and puts the content into its own words, allowing its users to simply copy and paste answers as their own. AI used this way greatly decreases the user’s ability to engage in the material, likewise understand and retain it and overall decreases their ability to learn about the questions they are asking for.

Future Considerations

With all of this in mind, I feel that the world is shifting towards a greater usage of AI. Many small and large companies began to implement AI into their applications, such as Discord, Notion, Microsoft, etc. and the many users that Chat GPT has garnered in these past years only expedite this developing field. I think that the widespread adoption of AI into day-to-day life could provide the general public the ability to learn more, yet unfortunately, I feel that this ‘advancement’ will succumb to only us becoming lazier and reliant on AI to do our work for us.

Conclusion

AI can work wonders in our daily lives, but it can also harm us when we apply it incorrectly. By intrinsically dissecting how we use AI to better ourselves and by being honest with our intentions with AI, we can all keep a fresh perspective on how we can use AI for self-improvement, rather than using to cheat us of learning content proficiently.