Skip to Main Content

Teaching with AI

How to Use AI Without Having it Write the Paper

Defeating the Blank Page

There are many ways that you can use Generative AI without actually having it generate the text of your paper. The most effective use in this case it to defeat the blank page. You can have Generative AI help with such things as:

  • Lists of elements of a paper
  • General outlines
  • Brainstorming topics or ideas
  • Reframing topics or research questions

Example:

Prompt: What are the required elements of a literature review?

  • Overview/Introduction: This should explain the focus and the purpose of the review and its scope. It should provide a brief explanation of the topic of interest.

  • Body: This is the main part of the literature review. It should have a logical structure and thematic divisions depending on your chosen topic. It typically includes:

a. Identification of the Works: This involves detailing the works you'll be analyzing for your review. It includes books, articles, or other sources that contain important and relevant information about the topic.

b. Summary of the Works: Here, you provide a concise explanation or summary of the major points and arguments in the works you are analyzing.

c. Critical Analysis: This is where you analyze the themes, structures, methodologies, and arguments in the different sources. You identify gaps in research, contradictions, controversies, strengths, and weaknesses.

  • Conclusion: This part summarizes the major findings from the review, discusses their implications, and suggests directions for future research.

  • References/Bibliography: This contains a list of all the sources that you consulted or cited in your literature review, according to the required citation style (e.g., APA, MLA, Chicago, etc.).

  • Inclusion of other elements such as context of the review, relevant theories or theoretical framework, and discussion of the research methods employed in the primary sources may also be necessary based on the requirements of the specific academic or research task.

Remember, a proper literature review should not just describe what the literature says, but provides a critical discussion on what has been published on the topic.

Getting Unstuck

Generative AI is also often useful for when you get stuck in research. You can also use Generative AI to suggest:

  • Search keywords
  • Synonyms for your existing keywords
  • Useful databases
  • Building library database searches

Example:

What are some useful for library database search expressions for a college-level paper on southern accents in popular media?

Certainly, here are robust search expressions using Boolean operators for your topic:

  • ("Southern Accents" AND "Media") NOT "Music"
  • ("Southern Accents" AND "Television") OR ("Southern Accents" AND "Movies")
  • ("Southern Accents" AND "Portrayal") AND "Popular Media"
  • ("Southern Accents" AND "Stereotypes") AND "Media"
  • "Media" AND ("Southern Accents" OR "Southern Dialects")
  • "Popular Media" AND ("Depiction" OR "Portrayal") And "Southern Accents"
  • "Linguistic" AND ("Representation" OR "Depiction") AND "Southern Accents"
  • ("Southern Accents" AND "Media") AND "Cultural Impact"
  • ("Southern Accents" AND "Popular Media") AND ("Perception" OR "Influence")
  • "Southern Accents" AND ("Film" OR "Movies" OR "Television") NOT "Radio".

These should help you better target your research for southern accents in popular media.

Note: You may have to specifically request that ChatGPT include Booleans in your prompt

Distilling Information

Caution: While the following is possible, be warned that AI may hallucinate or introduce other errors. It is up to you to ensure that the generative AI's interpretation is correct. 

Generative AI is also useful for distilling articles into more digestible forms. You can put text into Generative AI and ask it to: 

  • Create keywords based off of that article
  • Change the reading level of the article
  • Rephrase the article into a way that you understand. 

Example:

I need keywords for this abstract: Mahjong is a popular multi-player imperfect-information game developed in China in the late 19th-century, with some very challenging features for AI research...

Sanma, being a 3-player variant of the Japanese Riichi Mahjong, possesses unique characteristics including fewer tiles and, consequently, a more aggressive playing style. It is thus challenging and of great research interest in its own right, but has not yet been explored. In this paper, we present Meowjong, an AI for Sanma using deep reinforcement learning. We define an informative and compact 2-dimensional data structure for encoding the observable information in a Sanma game. We pre-train 5 convolutional neural networks (CNNs) for Sanma's 5 actions -- discard, Pon, Kan, Kita and Riichi, and enhance the major action's model, namely the discard model, via self-play reinforcement learning using the Monte Carlo policy gradient method. Meowjong's models achieve test accuracies comparable with AIs for 4-player Mahjong through supervised learning, and gain a significant further enhancement from reinforcement learning. Being the first ever AI in Sanma, we claim that Meowjong stands as a state-of-the-art in this game.

  • Mahjong
  • Sanma
  • Riichi Mahjong
  • AI
  • Deep reinforcement learning
  • Convolutional neural networks
  • Monte Carlo policy gradient method
  • Supervised learning
  • State-of-the-art

Abstract from: "Building a 3-Player Mahjong AI using Deep Reinforcement Learning" by Xiangyu Zhao, Sean B. Holden is licensed under CC BY 4.0

All responses taken from WTAMU's AI Hub, using ChatGPT4, build unknown.

AI Literature Review Tools

Warning for Off-Campus Users: None of these tools have integration with WTAMU Library's EZProxy, so please see the following FAQ to learn how to proxy the links provided by these tools. 

Many new tools have recently been developed that use generative AI's ability to look at the relationships between words to enhance the search for scientific literature. Many of these tools can be used to accelerate the literature Review process.

For more tools and information, please see TAMU Library's excellent guide on AI-Based literature review tools