Utilizing ChatGPT for Sensory Food Evaluation: Brownie Study
Click to start listening
Synopsis
A recent study indicates that ChatGPT can effectively assist in the sensory evaluation of foods, particularly brownies. It provides insights that may enhance product development and recipe optimization.
Key Takeaways
- ChatGPT aids in sensory evaluation of brownies.
- Study involves 15 different brownie recipes.
- Responses were categorized as positive, negative, or neutral.
- AI can streamline product testing processes.
- Future plans include refining ChatGPT for sensory evaluations.
New York, March 30 (NationPress) A recent study reveals that ChatGPT can be effectively employed in the sensory evaluation of foods, particularly brownies. This research provides valuable insights that could facilitate the development of new products and potentially improve existing recipes.
Researchers from the University of Illinois Urbana-Champaign in the US examined 15 distinct brownie recipes, ranging from conventional ingredient combinations to unique variations that incorporated mealworm powder and fish oil.
According to Damir Torrico, the assistant professor in the Department of Food Science and Human Nutrition, ChatGPT was given recipe formulas and tasked with describing the sensory attributes of each brownie, encompassing taste, texture, and overall enjoyment.
Torrico subsequently categorized the themes of ChatGPT’s replies to ascertain whether they were positive, negative, or neutral.
“Relying solely on human testers can sometimes impede the process, especially when numerous product prototypes require simultaneous evaluation. Sensory panels demand time and meticulous coordination, and in certain instances, some ingredients may not be food-grade, rendering them unsuitable for tasting,” Torrico mentioned in the study published in the journal Foods.
This is why large language models like ChatGPT are being explored for sensory evaluation, as they have the potential to mimic specific human responses.
Interestingly, ChatGPT’s feedback was predominantly positive, even for recipes featuring unconventional ingredients.
This favorable outlook aligns with the psychological concept known as “hedonic asymmetry”.
“Hedonic asymmetry” suggests that individuals (and seemingly AI) are inclined to describe beneficial experiences in more favorable terms. Food plays a crucial role in providing sustenance and energy, leading to a generally positive human response to food, Torrico explained, which ChatGPT appeared to emulate.
“Leveraging AI can yield general insights regarding which products warrant further testing and which should bypass the lengthy evaluation process,” Torrico stated. “I envision a future where ChatGPT is refined for sensory evaluation to support the industry.”
Looking ahead, Torrico aims to enhance the experiment by training ChatGPT to utilize a vocabulary akin to human descriptive panels.
