My opinion on AI:

At the moment, I do not think AI can be used academically starting from 300-level courses in Computer Science. Last time I used it was for a hackathon related to time series data. I am not that experienced in time series analysis so I have asked AI for suggestions on how to tackle time series data. It gave me some statistical tests to try, but they were not suitable. The problem is AI has assumptions about my data that are not true. I have discovered that after I read a book about time series analysis that discusses the very first step to even consider time series solutions to your data. That was the point that I knew I could not use AI to learn new things. 

On the other hand, it can be beneficial in summarizing notes, providing suggestions about topics or possible algorithms to tackle certain problems. It can be used for coding projects that have been done before. For example, AI can be beneficial in writing code that prints the output in a fancy way. Many tasks have been performed that tackled that issue and are public. Also, as a Math student, AI is useful for writing Latex, a markup language usually used to write scientific papers(try typst if you don’t, it compiles fast and it is easy to learn).

AI’s opinion on AI use in my field:

https://chatgpt.com/share/67152ed7-cfb8-8005-a11b-d5881f883b06

Summary of Interviews:

I have used AI to write the notes more concisely as I was in a rush while taking notes. I have also put additional info that I was not able to write but remember correctly:

Professor Mariam El Berry’s (Adjunct Professor at the Computer Science and Engineering Department at the AUC) notes can be summarized as follows:

  1. AI’s impact on performance: Large language models (LLMs) assist in organizing thoughts, offering starting ideas for projects, solutions, or even structuring research. They can be useful in studying by answering questions and providing aggregated knowledge.
  2. Academic use: While AI can quickly find references or help outline lectures, full reliance on AI can negatively affect a professor’s work. It is best used for starting points or hints, not for completing academic tasks entirely.
  3. Professional use: AI is more commonly used among professionals, especially software engineers. LLMs help provide examples of solutions to problems, such as Java code, saving time by allowing professionals to modify and test these examples.
  4. Classroom use: The professor does not use AI to structure lectures but does use it for graphical illustrations. Tools like Haiper.ai and ImagineArt are mentioned, though they need further improvement to be fully reliable.

Professor Ali Hadi’s (Professor at the Department of Math, Actuarial and Data Science at the AUC) notes can be summarized as follows:

  1. Use in student projects: AI can be helpful in student projects, potentially assisting in various tasks. It can be used to write end-to-end projects.
  2. Exam creation: AI could be used to write exams, particularly for generating true or false questions.
  3. Programming translations: There is potential to use AI for programming language translations. For example, if a person knows Python and wants to write code in R, it can be a good idea to use an AI.
  4. Personal experience: He once used AI in a task, but the result was poor, indicating limitations in certain creative or formal tasks.

Final Reflection:

I want to try ImageArt AI, I guess it would add a crucial part to my learning process as I am a visual learner. The only thing that came to my mind about AI was LLMs, but now, I need to consider other utilities as well. I was surprised that Dr Ali Hadi has not used AI that much. AI advancement in Latex can help you to write math formulas in a faster way. Nevertheless, Dr Hadi’s experience in the field makes him comfortable performing tasks without the use of AI. Another point I want to mention, my professor for the Real Analysis 1 course told me that she uses AI to solve a bug in her Latex code; however, that problem was not completely solved. This can be considered as an indication that AI is not that reliable even in easy tasks.