Does ChatGPT Stimulate Human Creativity?

BY Yuki Suzuki | May 24, 2023 | Student Writings

ChatGPT may not yet ease student learning at research institutions. ChatGPT has a tremendous advantage over humans in the amount of information it can memorize, and its responses sometimes contain professional details. However, you’ll be in a challenging situation if you do not have enough knowledge to paraphrase the information.

Conversely, the tool can be used as a check to review one’s lack of knowledge when information comes up that one does not know. As a senior student in an undergraduate program, I suspect that the more detailed the academic content, the more complicated it will be to study using only this tool since you need to check the information more carefully.

Fig. 1. ChatGPT answer

What is ChatGPT’s greatest strength? The responses that ChatGPT gives us are very straightforward. I enjoyed the fact that it responded with far fewer options than if I had asked the same question on Google. This reduction in options helps my willingness to learn more, even when I do not have time or my knowledge is limited. For example, when I asked the broad question, “Which art will help to understand Vietnam war?” ChatGPT offered five options from the following media: movies, novels, music, and memoirs (Fig.1). However, a Google search enumerated a list of questions to be asked that would make my question more specific (Fig.2).

Fig. 2. Google answer

When I ask a question about something I do not know, Google does not answer if the original question is unclear, so I need to rephrase the question in a way that is not misleading. On the other hand, the fact that the novels selected by ChatGPT overlapped with the ones I had read in the lectures I had taken proved that chatGPT has a good sense for what the user is looking for.

The possibility of more people wanting to learn, made possible by ChatGPT, would be superior to political information regulations that discourage people from learning. Moreover, if this tool brings people together who have knowledge and skill in many fields, there is even a possibility that a new art form could be born by mingling seemingly incongruous ideas. Therefore, my interest lies more in the reactions and expectations of those around me than in the ChatGPT system itself.

Looking to the past and present, information has become instantly available with the spread of the Internet, and I wonder if the current society expects everyone to have too much uniqueness. Unique perspectives take a long time to form because they are the accumulation of repeated small experiences. I am concerned that the emphasis on human creativity may be so great that the academic world will neglect to review basic ideas.

Basic knowledge is often treated as a formula in the world of physics and mathematics. How many students can firmly explain why a formula is correct? In history, it may be necessary to go back to what is written in textbooks. For example, many books say that the figure on the breastplate of the statue of Augustus at Prima Porta is Tiberius (Fig.3), but is he really Tiberius? My supervisor says this is not correct. Isn’t it a problem that attempting to reexamine the first part of the Roman Empire usually gets stopped in today’s educational institutions (in contrast to the study of excavations of new sites)? Reconsidering basic ideas should also be accepted as a form of creativity.

However, ChatGPT did not dramatically impact my perspective. This is because it was not new to me, as it produced all the same effects I have read in articles concerned with AI. For example, libraries’ online systems have been presenting sidebars of books that might be relevant for a long time, and this is an extension of that technological capability. Originally, such book recommendations would have been the job of the librarians.

Fig. 3. Copy of Augustus of Prima Porta gifted by Mussolini. Zaragoza, Spain.
Photo by the author.

In other words, it is still up to the individual to decide which services to use, and it is not desirable for society to over-regulate those services. I consider the emergence of ChatGPT to be a change within our control at this stage, although we must consider how we handle personal information and rights because the company behind ChatGPT may leak information to others without permission.

Finally, looking at the text structure in the ChatGPT responses, I see a tendency to separate complex events and summarize them by numbering them or using bullet points. These answers seem to be a classic form of writing. When we write a paper, we also formulate a question in the introduction that will be answered in the body of the article. We then analyze that question, develop a hypothesis, and repeat the analysis, which is the basis of our research methodology.

For students, a ChatGPT response can be viewed as an example of a thesis, even if it is not an academic one. Similarly, our papers usually include critical thoughts about past theories; our ideas are always in the process of being discarded and reimagined. Many organizations point out the potential dangers of ChatGPT, but many denials of new technologies have occurred in the past. At the same time, we can create a better future only with action, not thinking.

If ChatGPT’s responses stimulate our critical thinking and the joy of inquiry, we may find better solutions to environmental and other problems we face in the real world.