ChatGPT assists in brownie sensory evaluation, study suggests AI can aid food development

Researchers from the University of Illinois Urbana-Champaign in the US analyzed 15 different brownie recipes, incorporating both traditional ingredients and unconventional additions like mealworm powder and fish oil.

New York: ChatGPT can be used in the sensory evaluation of foods, specifically brownies, according to a study, which offers insights that could streamline development of new products, and possibly enhance recipes.

Researchers from the University of Illinois Urbana-Champaign in the US looked at 15 different brownie recipes, ranging from standard ingredient lists to unusual combinations, including mealworm powder and fish oil.

---

Study author Damir Torrico, assistant professor in the Department of Food Science and Human Nutrition, provided ChatGPT with recipe formulas, and asked the AI to describe sensory characteristics of each brownie, including taste, texture, and overall enjoyment.

Torrico then categorised the themes of ChatGPT’s responses to determine if they were positive, negative or neutral.

“Sometimes, relying on human testers can slow down the process, especially when multiple product prototypes need to be evaluated simultaneously. Sensory panels require time and careful coordination, and in some cases, certain ingredients may not be food-grade, making them unsuitable for tasting,” said Torrico in the study published in the journal Foods.

That is why large language models such as ChatGPT are being considered for sensory evaluation. It is possible to create models that can replicate certain human responses.

Surprisingly, ChatGPT’s responses were overwhelmingly positive, even for recipes that included unusual ingredients.

This positivity aligns with the psychological phenomenon scientists call “hedonic asymmetry”.

“Hedonic asymmetry” is the idea that people (and apparently AI) tend to describe things that are beneficial to them in more positive terms. Food serves a role in keeping us full and giving us energy, Torrico explained, causing humans to respond to food in a positive manner.

In ChatGPT’s attempt to act human, it seemed to display this behaviour. “ChatGPT was trying to always see the good side of things,” Torrico said.

“Using AI can give general insights of what products can be considered for further testing, and what products shouldn’t be put through that long process,” Torrico said. “I could see ChatGPT being developed for sensory evaluation to help the industry.”

Looking ahead, he plans to refine the experiment, training ChatGPT to respond with a vocabulary that is similar to a human descriptive panel.