What started in November 2022 as a general-purpose conversational chatbot, ChatGPT, has sparked a generative AI revolution whose impact continues to unfold. In a remarkably short span of time (especially when compared with past technological milestones like the invention of the printing press, the telegraph, or the telephone) generative AI has gone from novelty to necessity. Today, sectors ranging from education to space exploration are integrating this technology to boost efficiency and streamline processes.
Academic research is one of the areas I’ve been closely following and documenting. In fact, I’m currently finalizing a book on the use of AI in academic research, where I cover a wide array of tools, strategies, and practical tips to help researchers make the most of this emerging field.
That said, this post is about a very specific and powerful feature available in both ChatGPT and Claude (perhaps in other chatbots as well, though I haven’t verified yet). It’s called Projects, and in my view, it’s one of the most valuable but underutilized features for researchers.
If you’re using ChatGPT or Claude to support your academic work, and I believe you should be, then understanding how Projects works is essential. In what follows, I’ll explain what the Projects feature is, what it can do, and why it might change the way you approach your research.
What is Projects?
Projects is a dedicated workspace within ChatGPT and Claude where you can upload your research materials (PDFs, Word documents, slides, and so on) and interact with them using AI. Think of it as creating a custom version of ChatGPT or Claude, one that only draws from the sources you provide. All the answers, summaries, and insights you receive are grounded in those materials.
Currently, Projects is available for only paid users in both ChatGPT and Claude!
Why Does This Matter?
First, using Projects significantly reduces the chances of AI hallucinating. Since the chatbot is limited to a specific set of documents you’ve uploaded, it doesn’t pull information from its entire training dataset or attempt to generalize from the web. Its focus stays strictly on the sources you provide. This constraint increases accuracy and relevance.
Second, if you’re working on a literature review, which can be a massive task, having AI pull insights directly from your chosen papers is incredibly helpful. You can upload articles, ask targeted questions, and receive responses grounded in those texts.
Both ChatGPT and Claude currently cap Projects at 20 files, which I think is more than enough. In fact, as I mentioned in a previous piece about NotebookLM, the ideal number of documents to upload per project is somewhere between 10 and 15. Go beyond that, and you risk overloading the model, which usually leads to vaguer or less accurate outputs.
Now, when comparing Projects in ChatGPT and Claude with NotebookLM, I have to say I prefer the latter’s implementation. In NotebookLM, you can choose which specific documents you want to chat with from your collection. This flexibility makes the experience more focused and interactive.
In contrast, ChatGPT and Claude don’t let you select individual documents within a project to interact with. You can upload a single paper and chat with it directly but once you’re working inside a multi-document project, you can’t just toggle between papers.
There is a partial workaround: you can specify the paper you want the chatbot to focus on by referencing its title and author within your prompt. In my experience, though, this method is hit-or-miss. It doesn’t always yield accurate or consistent results.
So, if precision matters, and it usually does in research, I’d recommend uploading one paper at a time when you need focused interaction. This gives you cleaner, more reliable answers.
Ways to Use Projects in Your Research
Here are some practical ways Projects can help you in your academic research:
1. Reading the Literature
As I mentioned earlier, the Projects feature is incredibly helpful when you’re deep into your literature review. Instead of juggling multiple PDFs, highlighting bits here and there, and trying to keep mental track of recurring arguments, you can upload the papers into a single project and let the chatbot assist.
Ask it to identify key themes, summarize findings, compare methodologies, or even extract recurring arguments. From my experience, it’s like having a research assistant who’s read all your material and is ready to discuss it with you, on demand. It saves time and gives you a clearer view of the intellectual landscape you’re working within.
2. Cross-Referencing
Projects can also be a powerful tool when it comes to cross-referencing. If you’re trying to track who said what, or trace how a particular idea evolved across different papers, AI can help you map that out. With just a few targeted prompts, you can ask the chatbot to tell you which authors discussed a specific concept or how a term was used across studies.
This makes it easier to trace arguments, compare evidence, or spot subtle shifts in how topics are framed. It’s a much faster and more structured way of doing what used to take hours of back-and-forth flipping between papers.
3. Identifying Research Gaps
One of the most valuable uses of Projects, in my view, is its ability to help identify research gaps. Once you’ve uploaded your collection of papers, you can instruct the chatbot to analyze them for limitations, unanswered questions, or underexplored areas.
The model can sift through the discussions, conclusions, and future work sections to highlight where scholars themselves have pointed out gaps, or even infer where additional inquiry might be needed. This isn’t a replacement for critical thinking, of course, but it gives you a solid starting point, and often, some unexpected leads.
4. Refining Your Research Problem
Suppose you already have a research question in mind, but you’re not quite sure how original or relevant it is. This is where Projects can step in. You can upload a set of papers related to your topic and ask the chatbot to scan for how often the issue has been addressed, what angles have already been explored, and whether there’s room for a fresh take.
It helps you avoid redundant work and lets you fine-tune your focus. In some cases, it might even suggest more precise ways to frame your question or point you toward adjacent areas that offer more research potential.
5 Developing a Nuanced Understanding
One of the things I personally find most helpful is using Projects to deepen my understanding of complex concepts. Let’s say you’re trying to get a grip on how different scholars define and interpret a certain theoretical term, “digital literacy,” for instance.
You can ask the chatbot to extract all the definitions found in your uploaded papers, compare them, and explain the distinctions. You can also request examples or even ask it to simplify the language so it’s easier to digest.
6. Self-Assessment and Feedback
Another smart way to use the Projects feature is for self-assessment. Once you’ve read through a number of papers, you can ask the chatbot to quiz you on what you’ve learned. Instruct it to generate questions based on the content of the uploaded texts and use those as prompts to test your comprehension.
Even better, you can turn it into a conversation, ask the bot to challenge your understanding, push back on your answers, or guide you toward parts you might have overlooked. It’s a form of interactive feedback that helps reinforce your knowledge and highlight gaps.
7. Generating Counter-Arguments
You can also use Projects to help you think critically about your topic by generating counter-arguments. Once you’ve uploaded relevant literature, ask the chatbot to identify potential criticisms, opposing viewpoints, or alternative interpretations that challenge the main arguments presented.
This is especially helpful when you’re preparing to write a discussion or argument-driven section of your paper. It trains you to see beyond the surface of the text and consider different perspectives, something that’s often hard to do when you’re deep in agreement with what you’re reading.
Final thoughts
Projects is an important feature that every researcher or graduate student should seriously consider using. It offers a focused and efficient way to manage and interact with your reading material, allowing you to streamline your workflow at every stage, from reviewing literature and refining your research questions to developing deeper insights and critical perspectives. If you’re already using ChatGPT or Claude, this feature is right there at your fingertips, and in my view, it’s one of the most practical tools available for academic research today. Don’t overlook it.
