Fast ≠ Good: The Hidden Costs of AI Efficiency
ChatGPT's Deep Research represents a watershed moment in academic work. In mere minutes, it can produce a detailed, well-researched report that would have taken days or weeks to compile manually. While Deep Research can't replace true academic research, it does make aspects of it much more efficient. Although this dramatic increase in efficiency is impressive, it comes with costs that aren't immediately obvious. To understand these hidden trade-offs, we need to put this technological leap into perspective.
The Way We Were
When I started my doctoral program in the mid-1990s, we did have the ability to use online databases to find articles, but these were a shadow of what's available today. When I identified useful articles, I would go to the library to find the article in the bound collections of the journal. After scanning the article, I would either make extensive notes or, if the article was really important, go to the copy machine and make a copy of the article, page by page. This sounds silly now, but I still remember the joy I felt when the university rolled out a system that allowed us to use an ID card to pay for copies. It was inconvenient, time consuming, and expensive for a poor doctoral student. Now, it's a huge imposition to have to wait for inter-library loan to send a PDF of an article. The horror.
The Evolution of Research
In the 30 years since I started my doctoral program, the world of research has changed in some incredible ways. Google Scholar, for all its faults, makes research so much easier and more efficient, especially when coupled with library databases. Once you gain the necessary search skills, a few clicks and you have an article that you can annotate at will, share with your collaborators, store easily, etc.
AI is another big step in the evolution of research. Now, we can use tools like Research Rabbit, Elicit, and Cite to quickly identify useful articles, which we can upload and query in Google's Notebook LM. It's all a bit mind-boggling.
The Hidden Cost of Efficiency
But, this efficiency comes at a cost. Although the efficiency gains are wonderful, we've lost something in the process. For example, I would routinely find interesting articles when flipping through bound volumes of journals. These were articles that never would have been identified in a database search. Many times, they had nothing to do with the current project, but sparked ideas for later. This sort of serendipitous event is increasingly rare in our digital age. As we turn over more of the research process to AI, we'll lose other aspects of the human touch. Although I love the efficiency of modern research, a part of me mourns what we've lost.
There’s a broader point here. As I’ve mentioned before, there’s a huge danger in focusing on efficiency over effectiveness. AI has the potential to deliver both, but there’s a huge pull towards efficiency and while efficiency is important, we shouldn’t sacrifice effectiveness in the process.
The implications run even deeper than lost efficiency. When I stopped going to the library, I lost those moments of unexpected discovery that spark creativity. Novel insights rarely emerge from efficient, targeted searches. Instead, they arise when our minds wander and make unexpected connections. The most groundbreaking ideas often come from combining seemingly unrelated concepts—connections that algorithms, focused on relevance and efficiency, might never make.
The Creativity Connection
This is why I worry about students growing up as AI natives. Will they ever experience the productive randomness that leads to genuine innovation? There's a real risk that they'll miss out on the messy, meandering processes that foster deep learning and creative thinking. In our rush to embrace AI's efficiency, we might inadvertently sacrifice the very conditions that nurture human creativity.
Our challenge as educators is to find ways to preserve the serendipity and creative discovery that characterized traditional research while leveraging the powerful efficiency of AI tools. Perhaps the solution isn't choosing between old and new methods, but thoughtfully combining them to create something better than either approach alone.
Finding Balance
Should we resist AI and stop using it? Absolutely not. I’m all for using AI to gain efficiency and increase effectiveness. I just don’t want us to lose too much of the randomness and messiness of life that can lead to creativity and innovation. Just playing around, reading and thinking about random things can lead to great insights. Let’s not lose that in the rush to apply AI.
For those of you who teach, have some random conversations in class, include a few assignments that encourage exploration and process rather than outcomes. Encourage your students to let their minds go wild from time to time and, more importantly, to not always focus on tasks and productivity.
Let’s not forget that creativity and innovation are hard work that usually take unpredictable paths. Take some time to get away from AI and read random things, scan tables of contents, let your mind wander … whatever works for you. AI can amplify our creativity, but it can never fully replace the messy magic of human imagination.
Want to continue this conversation? I'd love to hear your thoughts on how you're using AI to develop critical thinking skills in your courses. Drop me a line at Craig@AIGoesToCollege.com. Be sure to check out the AI Goes to College podcast, which I co-host with Dr. Robert E. Crossler. It’s available at https://www.aigoestocollege.com/follow.
Looking for practical guidance on AI in higher education? I offer engaging workshops and talks—both remotely and in person—on using AI to enhance learning while preserving academic integrity. Email me to discuss bringing these insights to your institution, or feel free to share my contact information with your professional development team.