Language model-based Artificial Intelligence (AI) has become widespread through Chat GPT, an artificial system launched by Open AI. This chat-based web system is capable of performing tasks involving language, such as translation and text generation, as well as answering questions using information found online.
The system not only performs searches, but also generates coherent texts based on them. This has undoubtedly raised alerts in areas where text generation is part of daily activities, for example, among students who increasingly use this technology to avoid bibliographic searches that would otherwise take more time, or even to automatically generate essays. Teachers around the world are now concerned about how to detect whether schoolwork is actually done by students, and not automatically generated by AI. This is undoubtedly revolutionizing not only the way of evaluating, but also rethinking the ways in which, until recently, competencies and knowledge were accredited.
Beyond these uses, in the world of food, Chat GPT has also revolutionized everyday kitchen practices. More specifically, this AI is capable of generating cooking recipes or creating new recipes based on internet searches. It is not just a recipe search engine; it can generate dishes with special requirements. For example, it can be asked to create a recipe with ingredients found in the pantry or refrigerator—a great ally against food waste—or even modify existing recipes to obtain a vegan or gluten-free version. In everyday life, it seems to be a useful tool. However, the substitution of human ingenuity in the creation of new dishes by AI is one of the points of debate in the gastronomic world. Gastronomic study groups have taken on the task of investigating whether recipes for new dishes generated by AI from thousands or millions of existing recipes online are actually well-executed dishes. So far, it has been found that recipes for new dishes do not replace the human creative process, since some recipes generated by this AI were not very appetizing in their organoleptic properties.
All this is because, for the creation of new recipes, although the analysis of existing dishes could be a useful tool, technical and even chemical knowledge is also required to determine the technique of a dish, in addition to the subjective appreciation of the combination of flavors, ingredients, and textures generated by the senses. In the end, this AI is a powerful tool based purely on language, but if articulated with other data sources, it can be a great ally in everyday life for home cooks, or for developing new and better products for consumers, in the context of gastronomic industries that depend heavily on the foundations with which they develop better experiences for consumers—not only based on their needs and expectations, but also on their subjective appreciations.
— This article was originally published in Spanish by Liliana Martínez Lomelí. Translation generated with AI from the original text.
Read in El Economista →