I believe that my use of AI in my life will increase overall after taking this course, though I believe the uses will be very nuanced. The uses we had learned to use AI in class has the opportunity to become very reliable. This may not be the case currently as a lot of training has to be done for LLM’s to become free of hallucinations. In most cases it seems like if you need to fact check an AI’s response, which you should be anyway, it would be easier to confirm the research on your own.

As to how I would use, or not use, this in the future, AI will not be used for summarizing literature I have not read yet. I do not have enough confidence in a summarized piece of work done by AI to incorporate every key point in the text. I think AI could be useful summarizing writing that you have a decent understanding of already. This would serve as a quick way to remind yourself on something that may have been forgotten. Another simple use for AI is to help get ideas on how to organize drafts and or create other lists. These simple tasks would serve as a way to start something that would otherwise longer to do without AI.

After reading Tang et al., 2023, it made me think why not make it mandatory to inform the reader that AI was used in the production of the nursing journals? Any use of AI in an academic setting has the chance to be wrong. In a nursing setting, you would want to be certain the information you are publishing is correct. If a publisher decides to use AI in their journal and not inform the reader, the information could be trusted when it should not. This leaves a lot of liability on the journalist if something were to go awry. Surely, a journal article is not enforcing someone’s life decisions, but the idea that papers/generated videos can lead people in the wrong thought/direction is a concern. I will continue to talk more about this concept with my final project of Gen AI content/Deep Fake and how it can cause mass hysteria. I believe the solution to this is to disclose the use of all AI content to the viewer in every aspect.

In last weeks lab, the class went on to use AI in ways to engage in creative thinking. This was done by using AI to generate a poem as a start. It was very generic due to the weak prompting at first but after giving it some more specific details about what kind of poem AI was supposed to write, the LLM started to get more sophisticated with the response. This applied to other applications such as building a DnD character in my case. I have not played DnD before, but I am aware of the creative freedom the journey’s can go on, so I wanted to see what ChatGPT would create for me. I asked for a character name, backstory, and suggested items to ChatGPT.

When it comes to this creative response, I have to say I do not agree that something needs to be novel in order for it the be creative. A lot of this response has pieces in it that are, on their own, not original. Though, when you consider all of the pieces put together to make this unique character. That’s when you get novelty, surprise, and value all in one (Arriagada & Arriagada-Brunea, 2022). The surprise factor comes from the character’s backstory alongside the value. More value can be expected if you intend on using this character during a campaign. That is why I thought this response was great for creative outlook. DnD is full of creativity, and you could seemingly regenerate the same prompt and get endless amounts of characters to choose from. This method of using AI in a field that is already very creative is a perfect use for it. Creating characters in this way can cut down on time thinking, or even completely, while creating a unique character to go on campaigns.

After reading a paper in class regarding different ways to prompt AI, it allowed for more creative outlook on how to use AI (White et al., 2023). The most used type of prompting if I recall was in the form of persona. This is where you ask AI to impersonate whatever you would like. This is where creativity comes into it. I had Gemini (2.0 Flash) pretend it was an interviewer to try and see what kind of generic questions would come my way when I get that far into the job search. The specific prompt was, “You are a interviewer and I am the person you are interested in hiring, what questions would you ask me that would most relate to the geology field?” trying to focus on the field that I would be applying to. I understand most jobs will ask more specific questions regarding to what the company does, but I felt this is a good starting place for generic thought provoking questions. This kind of prompting goes further than just interviews. This simulates any kind of important conversation someone might have in which they want to prepare for. The less likely you are to be caught off guard, the better.

The response was not exactly what I had expected in terms of a conversation between interviewer and interviewee, but there were really good points brought up in why Gemini asked specific questions. It started with a generic greeting, and went straight into a line of questioning. What I found really neat about these though are the fact Gemini would explain what type of questions they were and what purpose that question would achieve by answering.

Link to paper cited: https://app.perusall.com/courses/writing-with-ai-sp25-gmds-19907-01/white-et-al-prompt-pattern-catalogue?assignmentId=6H83biAZq4wnKiRbg&part=1

AI is becoming a lot more accurate than I originally thought. By the end of the decade, I sense AI will be able to recite anything on the spot with precision. During Thursday’s lab period in which the class used three different sources for AI to summarize without using a document, I was surprised at the results. The common guess among the class was that the AI would not be able to pull specific details about the articles without the use of a pdf to upload to the LLM. My group, which had a very popular paper (Group 2), were shocked to see the accuracy of the results ChatGPT gave us on the summary of the paper.

Sure, it helped when a document was uploaded to the server, but the majority of the information was provided without the use of a pdf upload. The other groups had some similarities but group 3 had a bit more hallucinations from AI as the source was not as popular. As these algorithms train over the years, I feel like every source will become incorporated into their algorithms, allowing for the use of a simple search without the use of a pdf to read from.

The issue with this is that we can not consider what is “known” by AI. This simply means that whatever we input to AI such as ChatGPT cannot be considered fully accurate at its current state. Until labs like ours on Thursday start to become concise summaries across several papers without the need of a pdf upload, it will be necessary to always double check the paper itself to check for accuracy. Even then, our paper was more accurate with a pdf attached to the AI though it is always good to cross check papers when summarized by AI, pdf or not.