Uses for Generative AI
If AI is allowed in your course, this page offers ideas for using generative AI effectively. As earlier modules covered, ethics and academic integrity come first: always verify AI-generated content, attribute sources, and make sure AI supports rather than replaces your own thinking and creativity.
The AI-Human Loop: Why Human Oversight Matters
The term "AI-human loop" is used across fields like computer science, engineering, and data science to stress the value of combining human expertise with automated systems. Whether you use AI for school, personal, or professional tasks, bring your own knowledge, experience, and judgment into the process. When working with AI, engage actively rather than passively accepting its output:
- Ask questions and request clarification. Do not take AI responses at face value; dig deeper.
- Revise AI-generated text. Edit and refine to match your voice, purpose, and accuracy.
- Check AI-generated numbers. Verify calculations, data trends, and possible bias.
- Double-check AI-generated sources and facts. Confirm they are real and credible.
AI Toolbox: Practical Applications
Explore the categories below to see how AI can assist different parts of your academic work. These tools are meant to enhance your process, not replace your oversight.
Writing assistance
Use AI for specific parts of your writing process, such as brainstorming, structuring ideas, adjusting tone, and improving clarity. Always remain the primary author of your work.
Creative assistance
AI tools can support graphic design, music, storytelling, and the performing arts. Keep in mind that many AI models were trained on creative work without permission, which raises real concerns for artists and creators.
- Canva – design tool with AI-assisted templates and presentations.
- Mitsua Diffusion One – an image generator trained only on public-domain and opt-in data, so no copyrighted images were used without permission.
Protect your own creative work
- Have I Been Trained? – check whether your artwork was used to train AI models.
- Glaze / WebGlaze – apply distortions that disrupt AI training while keeping art recognizable to people.
- Nightshade – subtly alters images to "poison" models that scrape them.
Study assistance
AI can generate practice questions, explain concepts, and help with notetaking, but always verify the information for accuracy.
- NotebookLM – a research notebook that answers questions grounded in sources you upload. Its Audio Overview feature (launched 2024) turns your readings into a roughly ten-minute AI "podcast" between two hosts, which can be a useful way to review on the go. It only speaks English and can still introduce errors, so treat it as a study aid, not a source.
- Otter.ai – AI transcription and notetaking for lectures and meetings.
- TurboLearn – generates study tools, quizzes, and flashcards.
- Goblin Tools – breaks tasks into steps and helps with time management, designed to support executive-function challenges.
AI for accessibility
Used well, AI can lower barriers and expand access to learning:
- Goblin Tools – turns overwhelming tasks into manageable steps and can adjust the tone of your writing.
- Read-aloud and text-to-speech – built into most browsers, Microsoft Word, and your phone, these can read course materials to you.
- Live captions and transcription (for example, Otter.ai or built-in operating-system captions) support deaf and hard-of-hearing users.
- Be My Eyes – pairs AI image description with volunteers to help blind and low-vision users identify objects and read text.
As always, make sure AI is allowed for your assignment, and that the tool is approved for your campus account, before using it.
AI for Research and Data Analysis
Research and information gathering
AI can help summarize research papers and distill complex ideas, but always verify credibility by checking the original sources.
Generative AI assistants
- Elicit – uses language models to summarize and synthesize academic papers (verify its summaries against the originals).
- Scite – citation analysis plus a generative "Assistant" that answers questions with cited evidence.
- NotebookLM – a research notebook that generates answers grounded in sources you upload.
AI-powered research tools (not generative)
- ResearchRabbit – maps how papers connect through citations and suggests related work. It surfaces sources but does not write or summarize them for you.
Data analysis and visualization
AI can help interpret data and create visualizations, but make sure you understand and verify the outputs.
Generative AI
- Microsoft Copilot in OneNote – helps organize and draft research and notes (the account-linked option at (YOUR INSTITUTION)).
- ImageFX – Google's tool for generating images.
AI-powered computation (not generative)
- Wolfram Alpha – a computational knowledge engine that calculates answers for math and science from curated data and formulas. It uses AI techniques but does not generate new content the way ChatGPT does.
Knowledge Check
Select an answer to see feedback. Each option explains why it is or is not correct.
Question 1 of 4
What does the "AI-human loop" mean for how you use AI?
Question 2 of 4
A tool is listed in this AI toolbox. What can you safely assume about it?
Question 3 of 4
A tool finds related studies by mapping the citations between papers. It surfaces sources for you but does not write anything itself. What is it?
Question 4 of 4
A tool takes an article you upload and writes a short summary of it in brand-new sentences. What is it?
Attribution: this page is adapted from "A Student Guide to Navigating College in the AI Era" by the American Association of Colleges and Universities and Elon University, licensed under CC BY-NC.
