Generating Predictions About New Food Mind-Sets Using AI: An Insights- Development Technology Based Upon the Mind Genomics Way of Thinking

This paper presents a new approach to predict the features of food fads. The approach uses artificial intelligence powered and directed by Mind Genomics. Mind Genomics is an emerging science which got its start from understanding how people make decisions and very specifically how they weigh the different aspects of information to arrive at a decision. The actual history of Mind Genomics represents the evolution of purely empirical investigation [1,2] into an investigation using AI as a tool to suggest questions and answers and, finally, a tool to synthesize questions and answers. The history of Mind Genomics provides an interesting background to this particular effort to synthesize mind-sets about nutrition. This paper focuses on the last part, the use of AI to synthesize mind-sets. One might think about different aspects of nutrition and ask people in whichever way one wishes to rate the importance of those different aspects. Then, move from asking the people to move from general ideas about nutrition to rating specific ideas. That is give the topic of nutrition some “meaning,” not simple general words which fail to create an impactful word picture.

Mind Genomics took this topic and said that rather than having respondents, survey takers, rate the importance of single factors of a topic, say nutrition, it might be more realistic to present the respondents with combinations of single ideas or messages about nutrition (socalled “elements”) and instruct the respondent to rate the combination as if they were rating a story about nutrition, or about the nutrition of a new product. After all, goes the thinking, people in the real world are confronted with these combinations of ideas, not with single topics. It is unusual to work with single ideas. It is especially difficult to create “new products” or “new services” from strong-performing single ideas of a general nature.

Researchers using Mind Genomics ended up experiencing something totally unexpected. Many people, especially those who were more advanced in their fields had difficulty coming up with questions and then answers to those questions. And so, for many people, especially the professionals, the task of developing ideas seemed so contrary to their disciplined thinking that they reported experiencing difficulties and even irritation at doing the task. People could do the task, but a “coach” was needed.

ChatGPT was introduced to the public with great fanfare about two years ago and almost immediately began to exert its influence on critical thinking in education [3,4]. It was this juncture a year ago in the middle of 2022 when ChatGPT was put into the Mind Genomics platform (BimiLeap.com), along with an easy-to-use Idea Coach. The use of Idea Coach allowed the researcher to develop a paragraph leading to a “behind-the-scenes” prompt. ChatGPT embodied in Idea Coach ends up producing the questions and later, answers to questions, in a matter of seconds. It was easy to re-run the request to Idea Coach, generating so-called iterations. The Mind Genomics platform did all of the work. The process was so simple that grade school students in the third grade were able to become researchers [5].

The next step, leading to this paper, was a happy, fortunate coincidence where, strictly without expectation, we gave Idea Coach the entire background of the topic and ended up with far deeper and exciting results. Could Idea Coach (AI) move beyond asking questions and answers about a topic or postulating the existence of mind-sets? This approach was done in the spirit of AI-directed help to understand the future. Specifically, could AI powered by Mind Genomics thinking go more deeply, to describe in detail a variety of new nutritional mind-sets? The remainder of this paper is devoted to what we learned when we did the exact study that was done, and the results obtained. In doing so, the approach follows the approach taken by futurists [6-8].

Table 1 shows the prompts given to the Mind Genomics Program, BimiLeap.com. Instead of requesting a set of questions and answers based upon a short paragraph (the “squib”), the researcher wrote a longer, but equally low-information squib shown in Table 1. As can be seen, the table starts with a very short introduction about the respondents and the postulation that there are mind-sets of trends and fads in nutrition.

Table 1 requests the name of the mind-set, the key message, slogans which emblemize what to avoid and to consume, the motivations, the grounding of the mind-set and the predictions. Note that these types of questions simply ask the AI to give answers. The questions are targeted. No specifics are presented to the AI.

Table 1:Radical Detoxification.

Table 2 begins with the first of these far-out mind-sets. This mind-set is the RADICAL DETOXIFICATION. The key message is obvious: Avoid processed foods and toxins to cleanse your body and improve overall health. Numbers 2 and 3 are slogans. Slogans are easier to use to convince people. Number 4 deals with motivationswhy are people doing it? Number 5 assumes a nutrition magazine, while numbers 6, 7, 8 and 9 talk about sources of information in which this is built. Finally, steps 10 to 12 are predictions.

Table 2:The Direct AI Description of Mind-Set, RADICAL DETOXIFICATION.

Table 3 shows a subsequent, deeper analysis of the mindset generated by AI. The Mind Genomics platform provides a set of analyses, which move more deeply, analyzing the synthesized mind-sets.

Table 3:Deeper Analysis About the (Hypothesized) Radical Detoxification Mind-Set.

The objective of this paper is to show how LLMs (Large Language Models) can help the researcher to understand the increasingly complex world of mind-sets through artificial intelligence. Using LLMs allows the researcher to specify the mind-sets, or to request that the LLM synthesize mind-sets. When the LLM synthesizes mind-sets, the instructions can be to synthesize a limited number of radically different ones, or simply to synthesize different mindsets. The outcome is a clear understanding of what might be a very confusing, very business landscape. The power of LLMs is such that these explorations can be done quickly and with various questions asked of these synthesized or simulated mind-sets. The objective of this paper is to demonstrate the potential of AI to identify new or emerging mind-sets at a level of detail that moves beyond general predictions but does not make concrete predictions of the science. The mind-sets in this paper make sense, are not surprising and probably represent mind-sets which can be already guessed about. The importance of the approach is to stimulate critical thinking by an easy-to-use platform (BimiLeap.com), the incorporation of simple to write prompts (Idea Coach) and short iteration times in the span of 10-30 seconds per iteration, as well as rapidly provided deep analyses. At the end, if one were to define the one key benefit of this AI-based approach, it would be critical thinking which makes the user and the AI companions on the road to learning.

The authors gratefully acknowledge the help of Vanessa Marie B. Arcenas for preparing this manuscript for publication.

1. Moskowitz HR (2012) ‘Mind genomics’: The experimental, inductive science of the ordinary and its application to aspects of food and feeding. Physiology & Behavior 107(4): 606-613.
2. Moskowitz HR, Gofman A, Beckley JH, Ashman H (2006) Founding a new science: Mind genomics. Journal of Sensory Studies, 21(3): 266-307.
3. Adeshola I, Adepoju AP (2023) The opportunities and challenges of ChatGPT in education. Interactive Learning Environments 1-14.
4. Fütterer T, Fischer C, Alekseeva A, Chen X, Tate T, et al. (2023) ChatGPT in education: Global reactions to AI innovations. Scientific Reports 13(1): 15310.
5. Mendoza C, Mendoza C, Deitel Y, Rappaport S, Moskowitz H, et al. (2023) Empowering young people to become researchers: what does it take to become a police officer? Psychology Journal Research Open 5(3): 1-12.
6. Avin S (2019) Exploring artificial intelligence futures. Journal of AI Humanities 2: 170-191.
7. Harfoush FA (2021) Predicting the future. In Management for Professionals pp. 6-80.
8. Mellers BA, Lu Y, John PM (2023) Predicting the future with humans and AI. Consumer Psychology Review 6(1): 109-120.