Why Use AI?
Numerous options are available for students to use AI ethically, while following the specific directives of their course professors. AI tools can be used for academic inquiry and to do research effectively and efficiently; however, caution must be exercised regarding AI-generated outputs for potential biases, methodologies, reliability, and integrity.
Leveraging AI for Academic Success
| Research & Topic Development | |
 |
Topic Brainstorming: Generate ideas, explore perspectives, challenge assumptions, create research starting points Topic Refinement:
Alternative Perspectives: Identify related concepts, consider alternative viewpoints, enhance critical thinking |
| Search & Research Optimization | |
 |
Keyword Development: Generate relevant search terms and subject headings, improve database searches, identify main themes Boolean Logic & Database Selection: Formulate precise searches, narrow results, identify appropriate databases (subscription-based and open access), focus research efforts |
| Learning & Comprehension Support | |
 |
Concept Clarification: Get detailed explanations or clarification of complex topics, receive feedback tailored to different learning styles, access examples and illustrations, facilitate comparative analysis between concepts Text Processing: Summarize documents, identify key arguments, get overviews, process data efficiently |
| Organization & Planning | |
 |
Structure Development: Create outlines, visualize ideas, identify gaps, plan research, clarify focus |
| Language & Cultural Support | |
 |
Non-Native Speaker Assistance: Improve writing/reading, reduce language anxiety, understand cultural references |
| Data Analysis & Insights | |
 |
Pattern Recognition: Analyze trends, make evidence-based decisions, interpret information faster |
| Expert Identifications | |
 |
Scholarly Networks: Find key experts, access diverse perspectives, stay current with research |
Risks of Over-Relying on AI
Over-Reliance and Dependency on AI Technology
Relying solely on AI tools for research can hinder the development of analytical skills needed for academic success and may also stifle students' creativity. When using AI for research, students should use caution and be aware of potential distortions and limitations, such as those listed below.
Hallucinations refer to instances where Large Language Models (LLMs), such as GhatGPT, generate information that is false or misleading. These may be affected by the user's prompt, if it is not critically thought through. Hallucinations may be inaccurate statistics, fabricated references, misinformation, or invented facts that look credible but are not based on facts.
Bias - When an AI model is built, the training data may contain skewed data, imbalanced information, or prejudiced examples. As a result, AI-generated responses can reinforce stereotypes or favor certain perspectives. Additionally, the user's prompts, preferences, or prior interactions can further influence the output, potentially introducing unintended bias.
"Tortured phrases" are unnatural or convoluted language produced by LLMs that obscures the original meaning, especially in technical terms or simple expressions. This often occurs when LLMs attempt to translate or paraphrase input without fully understanding the context or cultural nuances. For example, when words are translated between languages (e.g., from English into a foreign language, then back into English), the original meaning may be lost or distorted, resulting in an awkward and misleading "tortured phrase."
| Technical Terms | Tortured Phrases (Over-complicated Substitutes & Convoluted Replacements) |
|---|---|
| Big Data | • Massive-scale datasets • Gigantic information pools |
| Artificial Intelligence (AI) |
• Manufactured cognition |
| Breast Cancer | • Chest-related malignancy • Bosom affliction |
| Climate Change |
• Global thermal shift |
| Heart Disease | Cardiac malfunction ailment Coronary vessel malfunction disorder |
Retracted Works - Does the AI response include information from retracted articles? Since AI tools may not have been trained to recognize and exclude retracted articles, using information or data from an article that has been retracted may result in flawed research outcomes.
Mind Map: How GPTs Outlines & Visualizations of Ideas Help Students
Note: For a mind map of your GPT 4.o text, use a simple prompt: "create a mind map" or "create a mind map of above."
Selected References
Cabanac, G., Labbé, C., & Magazinov, A. (2021). Tortured phrases: A dubious writing style emerging in science. Evidence of critical issues affecting established journals. arXiv:2017.06751. opens new windowhttps://doi.org/10.48550/arXiv.2107.06751
Osama, M., Afridi, S., & Maaz, M. (2023). ChatGPT: Transcending language limitations in scientific research using artificial intelligence. Journal of College Physicians and Surgeons Pakistan, 33(10), 1198-1200. opens new windowhttps://www.doi.org/10.29271/jcpsp.2023.10.1198