At my high school, seniors had to present a research project on any topic of their choosing. For my presentation, I decided to focus on the global impacts of the fast fashion industry, and I studied everything about the topic for five months. Through my research, I learned about the different fashion production models, the global fashion supply chains, the overconsumption of apparel, the related environmental and ethical issues, the solutions to these issues, and so much more. Therefore, when I was prompted to challenge ChatGPT’s knowledge of something I knew a lot about, I knew exactly what question I would ask.
I asked the AI to define fast fashion, and according to ChatGPT, “Fast fashion is a business model and a term used to describe the rapid production of inexpensive, trendy clothing and accessories that are designed to quickly respond to the latest fashion trends.” This was a solid business definition of the term “fast fashion.” At the beginning of my presentation, I referred to a similar definition of the term. It also listed and explained six defining characteristics of fast fashion: quick turnaround, trend imitation, low prices, high volume production, short product lifecycle, and global supply chains. These characteristics seemed very familiar to me, and then I realized that each of these characteristics was a topic I had discussed in my presentation!
The explanations for these characteristics were correct but vague, and they could be easily misinterpreted. For example, ChatGPT’s explanation of high-volume production was, “Fast fashion retailers produce clothing in large quantities to meet the high demand for trendy items.” This is technically true, fast fashion companies do produce a lot of clothes, but the traditional fashion business model does as well. What ChatGPT fails to specify is that fast fashion is unique because it produces a wide variety of clothing items, and produces those items in smaller quantities. In contrast, traditional fashion does the opposite of this: they will design fewer clothing options and produce them in large quantities.
I wanted to see if I could get ChatGPT to give me the right answer, so I asked it to tell me more about the high production volume of fast fashion brands. It gave me an answer in the exact same format: a paragraph, then a numbered list, followed by a concluding paragraph. This list gave me information about fast fashion’s use of economies of scale, assembly lines, frequent inventory turnover, just-in-time production, fast supply chains, and trend analysis. Though some of these terms, such as just-in-time production, would support my answer, none of the information in the list provided me with the conclusion I was looking for.
So, I asked my third and final question: “How are the production cycles of fast fashion companies different than traditional fashion companies?” Finally, in the same format as before, ChatGPT gave me my answer. It explained that fast fashion retailers maintain a low inventory volume and higher production quantities, but traditional fashion retailers have a large inventory volume with a low production quantity.
This conversation with ChatGPT taught me that AI might not always be an accurate source of information. The responses were always vague and repetitive like it was trying to say everything it could in hopes that something it said made sense. It also seemed that the more specific the question became, the less confident ChatGPT responded. Therefore, we should still rely on our research and not let AI do the research for us.
However, I do believe that the chatbot is very helpful in showing us what to research. It took me weeks during my research to discover the defining characteristics of fast fashion, and it took me even longer to find articles that related to those topics. In Kotorchevikj’s article on the use of ChatGPT in research, she says, “You can use AI to craft your research sections and main points that are still unique and original.” If I had asked ChatGPT to come up with the main points relevant to my research of the fast fashion industry, I could have saved myself two weeks of work. If used correctly, ChatGPT could be an incredible tool to help speed up tedious work in the research process.
Works Cited
Kotorchevikj, Ivana. “Elevate your Research: Best ChatGPT Prompts for Research.” Neutral Network Press, 2023, neuralnetworkpress.com/best-chatgpt-prompts-for-research.