Before you start using AI to support your work, consider the following:

1. What output are you looking for? Determine whether the tool you plan to use can do what you want. See the "Types of tools" page for ideas.
2. How will you use AI ethically? AI can be an excellent support for brainstorming and enhancing understanding, but your professors and the places you might publish are likely to have guidelines you should follow. Consult them before incorporating AI into your work.
3. Do you have time to evaluate the output properly? AI tools can generate incorrect information; if asked, they may declare that the information they've produced is correct (so you can't take the tool's word for the veracity of the output!). Locate other reputable sources to confirm any claims before proceeding with your work. The SIFT technique can be helpful here: it asks you to Stop, Investigate the source, Find better coverage, and Trace claims to their original source. Learn more about the SIFT technique below.

Using well-crafted prompts is critical to getting the most from AI tools. Here are some tips to help you develop effective prompts.

Prompt using clear and straightforward language

• It's not necessary to include jargon or complex wording. Use prompts that are conversational to start, and add complexity if needed.

Provide context

• What kind of output are you looking for? What level would you like it to be at? Include this information in your prompt.
• Sample: I'm writing a college-level paper about social media and mental health for my advanced psychology class. What are some topics I could research?

Add synonyms or details to fine-tune results

• If you're not happy with the results provided, add more information to get a more helpful response.
• Sample follow-up: Provide topics more focused on college students or young adults.

Regenerate results and provide feedback to the AI tool

• Many tools offer a "rewrite" function so that a new response is created, as well as a "like" or "dislike" function to provide feedback to the programmers. This helps improve future output.

If you're using AI tools to support your research, learning, and academic work, keep in mind generative AI's strengths and weaknesses and tailor your use appropriately. 

Good for...

• Getting quick background information on a new topic
  • In other words, as a starting point for inquiry rather than something you consult later in the research process
• Brainstorming in preparation for an assignment or overcoming writer's block
  • If you're stuck in your research or writing process, you can use generative AI as a sounding board to bounce ideas, using iterative, conversational prompting to develop your own thinking (kind of like talking to a peer or a tutor)
• Considering a topic from new angles or perspectives
  • Generative AI can be an excellent way to develop a more nuanced understanding of a topic you're interested in or come up with new and exciting dimensions to your research you hadn't previously considered
  • In other words, it can be an inspiration for taking your own work in new directions!
• Outlining or organizing your ideas
  • Use generative AI as an initial step in your writing process to test out different ways of structuring your arguments and organizing your writing
• Repackaging your work
  • Need to transform your work from narrative text to a table? Generative AI can be a great time-saver if you need to tweak the format or presentation of text you've already crafted

Use caution when...

• Asking for a list of sources for research (you may end up with a bunch of sources that sound great, but don't really exist)
• Generating a list of prominent researchers or thinkers in a particular field (you may get output with a gender or racial bias, or you might get a list of people that aren't real!)
• Asking for high-level information you don't have the expertise to evaluate (because AI tools can hallucinate, generative AI is a great tool for background information, but you should only ask it questions you have the capacity to fact-check yourself)