Artificial Intelligence Tools for Use in Academic Contexts
What is the Focus of This Site?
Generative artificial intelligence (AI) has grown increasingly popular over the last several years. Relying on large language models that can be trained to recognize, generate, and perform a variety of tasks with text, generative AI is used in a variety of digital tools such as ChatGPT. Its growing appeal stems from its ability to learn from its interactions with users and use their inputs to produce new information with similar characteristics.
The purpose of this site is to familiarize students, faculty, and staff at Saint Catherine University with popular generative AI tools and their applications. It offers guidance on using these tools, and points out additional resources that can be used to learn more about them. It also elaborates on some of the ethical considerations of using AI while laying out basic citation rules.
What are the Basics of Generative AI?
Generative AI can be helpful with many tasks, including:
- Drafting outlines for writing
- Suggesting synonyms in place of other vocabulary in papers
- Copy editing
- Brainstorming ideas for an essay or project
- Summarizing the content of an article
- Translating the content of written materials
Generative AI tools tend to be less effective in assisting individuals with the following tasks:
- Inquiries that require previous knowledge of a topic
- Searches for very recent information
- The computation of complex mathematical calculations
- The creation of something that requires original thought
How do I craft quality prompts for AI?
An AI prompt is a command or question that a person inputs into an AI tool to get it to manufacture a result. Prompts can be text, information, coding, or questions. The AI model analyzes the prompt and generates a response based on what it has learned. A well-crafted prompt enables the AI to give you a response that is relevant to your input, meaningful, and utile. A bad prompt may result in irrelevant data or direct you to suboptimal research. Overall, good prompts are concise, logical, explicit, adaptive, and reflective.
In general, the effectiveness of an AI prompt depends on the way in which it is structured and the extent to which it adequately captures the intent of the user. According to the Harvard University Information Technology team (2023), strategies you can use to write an effective query or prompt include:
- Being specific: Generic prompts like “Write a poem” will produce generic results. What kind of poem do you want? What is its style? Who do you intend to read the poem? How long should it be? Should it be happy or sad? Adding this type of context and being as detailed with your input as possible will help generate more useful outputs.
- Asking the AI to Assume a Role: An effective way to get generative AI to produce meaningful and useful responses is to ask it to behave as if it were a type of person, process, or object. The AI will attempt to assume the role it was given and tailor its answers accordingly. For example, if you are drafting an argumentative essay about the risks and benefits of a surgical intervention, ask the AI to "draft an outline as if you are a doctor."
- Telling the AI How to Present Its Response: Generative AI can produce many different types of outputs, including HTML or CSS code, fictional tales, research summaries, dialogue, revised emails, podcast transcripts, images, and more. Specifying the type of output you want from the AI tool you are using will yield better results. After describing what you want, try adding “Present this in the form of…” and your preferred output. Or start your prompt with, “Create a [format of output] about / that contains…” etc.
- Identifying What You Want Shared: Clarifying to AI what specific details you would like it to include and exclude from its response can make your interaction with it more efficient and improve your result. For example, if you are interested in finding articles about farming practices of specific indigenous groups in the United States, tell the AI: "Do include the Ojibwe and Dakota. Don't include the Chippewa and Ho-Chunk."
- Providing the AI With An Example: An AI tool is more likely to give you a useful response if you provide it with an example of a sentence or paragraph that you would like your output to resemble. With that said, it is very important to abide by copyright law when interacting with AI. Never upload copyrighted works to AI and ask it to rewrite or reproduce it or create a similar work.
- Specifying the Use of the Information You Are Seeking: Provide AI with as much detail as possible about who you anticipate will read or respond to your work, and be sure to specify the tone you would like the output to embody. For example, "Give me ideas for a graduation speech that is inspiring and comical but appropriate for youth, adults, and elders" will generate better results than typing "Write a graduation speech."
- Building on Previous Prompts: It is not essential, nor always effective, to include every imaginable detail in your first input to AI. In your first prompt, begin with a basic question and then add details to your subsequent inputs. Change the wording or tone or add more context and specificity to guide the AI toward the output you hope it will produce.
- Giving the AI Feedback: Pretend you are collaborating on a project with AI, and speak with it as if it were a colleague or teammate. Do not be afraid to give it feedback – let it know which parts of its response were useful and which parts could be improved. If you notice it got something wrong, tell it so it can correct its mistake.
- Asking the AI What It Needs to Know to Help You: Having trouble with creating a good first prompt? Ask AI to help by manufacturing one for you! Begin with a simple description of what you want and ask the AI to elaborate upon it for you, like “What do I need to share with you to help me locate articles on intergenerational programming in childcare settings?”. And simply adding “Tell me what else you need to do this” at the end of any prompt can fill in any gaps that will help the AI produce better outputs.
Where can I learn more about AI?
Ithaka S+R has created a Generative AI Product Tracker which lists tools by their primary purposes and includes pricing information, as well as updates on the tools' features and limitations.
References
Georgetown University Library. (2024, November 21). Artificial intelligence (generative) resources: How to craft prompts. Click this link to view this resource
Harvard University Information Technology Department. (2023, August 30). Getting started with prompts for text-based Generative AI tools. Click this link to view this resource
Last updated December 4, 2024
Back to top