Artificial Intelligence: Is it A Useful Tool?
07 May 2024My Experience With Using Artifical Intelligence
I. Introduction:
In my Software Engineering course, AI emerged as a valuable resource to aid us in completing our work assignments. AI, short for Artificial Intelligence, serves as a versatile tool enabling individuals to seek assistance in various tasks. It allows users to pose questions and receive answers across a wide range of topics. Notable AI platforms like ChatGPT and MicrosoftCoPilot exemplify this capability. Throughout the course, we explored the multifaceted applications of AI, leveraging its potential to streamline tasks, provide insights, and enhance problem-solving. This introduction sets the stage for delving deeper into the challenges, opportunities, and considerations surrounding the integration of AI in our coursework and beyond.
II. Personal Experience with AI:
I have used AI in class this semester in the following areas:
- 1. Experience WODs e.g. E18
- For the Experience WODs, I only followed the videos after I did the code myself since, I thought it was the best way to learn the method.
- 2. In-class Practice WODs
- For practice WODS I follow the same thought process similarly to Experience WODs, to not use AI.
- 3. In-class WODs
- For In-class WODs, I used AI to assit me since I was confident enough in my own coding skill in which I could use AI to help me by setting up the code template.
- 4. Essays
- I only used ChatGPT to correct any grammar and spelling errors since the software doesn't have a built in auto-correct.
- 5. Final project
- I tried using AI for this but it didn't seem to help and gave more errors than solving them.
- 6. Learning a concept / tutorial
- I used AI by asking for examples on how to use methods like the java _.map method, so I would ask questions like "Can you give me an example on how to use the _.map method" and then apply it to my code.
- 7. Answering a question in class or in Discord
- When answering a question I would just copy and paste the question into ChatGPT, but I would keep in mind that the answer wouldn't be 100% correct so I would then research it bymself to check if it was accurate enough.
- 8. Asking or answering a smart-question
- For answers a smart-question, I wouldn't consult AI for this since, in my experience AI has a hard time answering questions that have depth or it would have a hard explaining it after answering.
- 9. Coding example e.g. “give an example of using Underscore .pluck”
- Asking AI for coding examples like "Give an example of using _.map", "Give an example on how to set up CSS", "Give a template with these method names...".
- 10. Explaining code
- I would paste in code I find from the experience WODs and ask the AI for further explaination after looking at the video.
- 11. Writing code
- For asking AI for code, I only ask it for simple tasks, since the more difficult the task, the more of a chance of it having errors. So, usually I don't ask AI for this.
- 12. Documenting code
- I usualy don't use AI for Documenting code, however I do ask about potential test cases that I might need for the function.
- 13. Quality assurance
- I think the quality differs based on which AI you are using however, most AI is good enough for some things just not all.
- 14. Other uses in ICS 314 not listed above
- Because I started using AI in ICS 314, I've been using it to point out grammar mistakes on my essays for different classes.
III. Impact on Learning and Understanding:
Using AI can either enhance or diminish comprehension, skill development in respective subjects, and problem-solving abilities. One way AI could weaken problem-solving skills is by relying on it as a shortcut, akin to cheating. Depending solely on AI to solve problems deprives users of the chance to learn independently and improve their own skill sets. Similarly, when AI solves problems entirely, users miss out on the opportunity to develop their problem-solving skills as the AI performs all the work, hindering personal growth. Conversely, AI can aid skill development by providing error corrections and guidance on coding methods. Seeking AI assistance for error corrections and coding methods offers users additional examples to learn from and apply to their work. For example, requesting AI assistance in error correction allows users to learn from their mistakes, reducing the likelihood of repeating them. Furthermore, AI can enhance learning by suggesting alternative approaches to problem-solving without solving the problem outright. Instead of asking AI to complete a task, users can request examples of diverse methods for accomplishing their projects. This approach encourages exploration and deepens understanding of various problem-solving strategies, enabling users to expand their skill sets.
IV. Practical Applications:
I'm aware of several examples of AI being utilized in various projects, including games and machine learning. Online videos often demonstrate content creators developing Machine Learning Algorithms (MLAs) to achieve specific objectives. For instance, a [YouTuber](https://www.youtube.com/watch?v=DmQ4Dqxs0HI) recreated a game in code and design an AI to play and succeed in the game while adjusting parameters and achievements for the AI. Another example involves a [YouTuber](https://www.youtube.com/watch?v=pJPdW8WWAso) creating an MLA that generates a 3D character model and instructs it to walk. These projects are intriguing as they illustrate the progression of AI and highlight the need for continuous evolution to accomplish tasks effectively. They demonstrate the effectiveness of MLAs in tackling simple tasks given sufficient time, minimal requirements, and ease of implementation. A real-world application of AI in practical work involves an MLA scanner that was initially developed for scanning bread to calculate purchase costs. However, it was later repurposed for identifying cancer cells due to its capability in recognizing different shapes and sizes and comparing them. You can find more about it in this [New Yorker article](https://www.newyorker.com/tech/annals-of-technology/the-pastry-ai-that-learned-to-fight-cancer).
Challenges and Opportunities:
Some challenges that arise when using AI include using any work or code provided by AI, which may eventually lead to errors. Relying solely on AI to complete an assignment is risky and detrimental to the worker, the consumer, and the project as a whole. Employing code generated by AI is a significant mistake, as it may contain unforeseen errors. Examples include errors that contradict or interfere with other files. If the code needs to be changed, users who have copied the code may struggle to update it since they did not write the code themselves. However, there are opportunities when using AI. These include saving time on work, leveraging MLAs to accomplish tasks, and seeking AI assistance in error correction. These avenues offer efficient ways to incorporate AI into projects. For instance, requesting AI to format a project template saves time that would otherwise be spent on manual formatting, allowing users to focus on other aspects of the project, such as selecting methods and functions. Similarly, employing MLAs to accomplish tasks, as mentioned in the previous section, saves time in the long run and can be repurposed for different areas if used correctly. Additionally, utilizing AI to fix errors is beneficial, especially when debugging code can be challenging without recognizing the error line.
VI. Comparative Analysis:
Comparing the teaching methods of a standard face-to-face teacher and an AI, the standard teacher is by far more advanced, especially in coding contexts. When consulting a professor or senior regarding coding, they provide a multitude of examples and various approaches to problem-solving. In contrast, AI typically offers a singular approach, deducing it as the best method for the task. While AI may provide efficient solutions, learning from a professor offers more diverse examples and guides learners through the coding process with different methods. Overall, engagement levels are higher with a professor, as questioning is more thoughtful. Knowledge retention is also higher with a professor, as mental notes tend to be more effective than merely copying and pasting code provided by AI without understanding its implications. Additionally, learning under a teacher allows for more hands-on practice, fostering skill development in coding.
VII. Future Considerations:
When utilizing Artificial Intelligence, several factors merit consideration. Firstly, we must explore the various ways to integrate AI into our work and daily lives, leveraging its capabilities to streamline tasks and enhance efficiency. Additionally, we should explore how AI can aid in self-education and teaching others. For those contemplating future AI use, it's advisable not to rely solely on AI when learning something new. Instead, individuals should conduct their own research through articles, online courses, or instructional videos. AI should serve as a supplementary resource, particularly when learning coding methodologies. Furthermore, it's essential to recognize areas where AI requires improvement. AI systems may not always be up-to-date with the latest information, as their knowledge is limited by their creators. For instance, ChatGPT's knowledge is current only up to January 2022. Thus, if developments occur in coding or related fields after this period, AI may lack updated information.
VIII. Conclusion:
Is AI a useful tool? Yes, undoubtedly, AI proves beneficial for specific tasks within work environments. However, relying solely on AI for entire projects offers limited benefits. AI should augment human efforts, not replace them entirely. Reflecting on my own experience with AI, I initially found it helpful in this course. Yet, I realized it wasn't facilitating true learning; it was merely aiding in completing tasks. Over time, I adapted my approach to using ChatGPT's AI to my advantage, focusing on learning rather than simply copying and pasting code. In the realm of coding, AI should be utilized primarily for creating templates, error correction, providing additional examples, and serving as a last resort when encountering unfamiliar concepts.