AI & Academic Research
How should AI be used in the research process?
First, you should consult whatever the boundaries set by your professor for your particular class and assignment are. Some faculty may have policies against any AI use during the research or writing process.
Secondly, you should think about your own boundaries and ethics of AI use. Modern academic was established well before the advent of AI, so it’s certainly not strictly necessary for you to engage with to be a successful academic researcher and writer. However, many students will be thinking about how they might use AI in their future workplaces, and of course many folks will want to use it to help with research.
Just like someone else can’t work out and build your body for you, academic work is intended to build your actual brain muscle/connections in the form of research, critical thought, analysis, expression/communication, revision and editing. The most successful folks will be those who use AI as a tool to be more creative and innovative, not as a way to avoid the mental work that builds critical thinking skills!
When it comes to searching for resources, ideally AI, if you’re using it at all, should be just one search tool you are using alongside traditional web search strategies like searching databases and even the open web for key phrases and subjects.
AI Strengths and Weaknesses
In general, AI is good for:
- brainstorming
- exploring foundational concepts
- giving an overall narrative of a discipline versus you gathering sources and assembling your own
As of 2025, it is still pretty bad at:
- accurate citations
- reliably factual information, in fact it will garble accurate information given to it sometimes in unexpected or hard to spot ways
- copyright
- original content
How can AI help with new research concepts and terminology?
One notable strength of current generative AI is the ability to explain or summarize new content. Article abstract full of a bunch of words you don’t know? Ask ChatGPT to summarize it for you!
AI Challenge
- Ask your preferred AI about a topic you know well, and see how accurate the information is.
- Now ask it about something you know nothing about, how would you go about verifying the information of the topic you don’t know? Spoiler alert, it is hard to do!
What’s your strategy for how you can use AI for what it’s good at while minimizing the impacts of its weak points?
There’s no one right answer, but your strategy should use AI for what it’s good at (general searching and concept formation) and avoiding relying on it for areas where it struggles (factual data replication, citations, writing content.).