What do you bring to Thanksgiving when your aunt is gluten-free, grandma should avoid sodium, and your brother has a dairy allergy? Finding a dish that can meet everyone’s dietary restrictions while still being delicious is a challenge.
Sony AI believes that in the near future,can help.
Sony AI pursues a digital web that enables rich conversations between AI systems and chefs, helping them take their creations even further and develop new and unique recipes. Their research in the field of gastronomy aims to one day change the way chefs create food combinations, pairings and plates and to help chefs in their process of developing new, original, equally healthy and eco-friendly recipes. environment. This includes matching ingredients and creating recipes.
This is an interesting use case for AI, given the position of gastronomy, straddling science and art. To learn more about the effort, I reached out to Dr Michael Spranger, COO of Sony AI, to discuss the methodology of AI technology and future applications in the world of gastronomy.
GN: How did Sony AI first become interested in food as a technology use case?
Dr. Michael Spranger: Simply put, gastronomy is a fascinating and unexplored field for Sony. Just like music, movies and games, we see food as a global entertainment company, whose technological advancements constantly contribute to its progress. Over the years, kitchens have evolved with new technologies, and chefs bring the purest creativity to this technology.
Our goal is to develop technology, and specifically AI and robotics, for chefs that allows them to be even more creative in crafting delicious dishes while helping them deal with issues such as health and durability.
In the realm of recipe creation, we believe AI can help chefs explore vast amounts of data associated with foods, including existing recipes, chemical and molecular makeup, and other factors such as data. nutritional or environmental impact to create a new dish. With robotics in the kitchen, we hope to help chefs with their cooking process, from preparing and cooking food to serving and serving. None of these research areas are easy to solve, and that’s exactly why we thought they were appropriate for us to define a big challenge in our flagship gastronomy project.
GN: There is a perception that AI and automation are replacing humans. How do you see chefs working with technology, and what has the reaction been so far?
Dr. Michael Spranger: Sony AI’s initial focus is on high-end restaurants and their chefs. Our role is to improve what chefs are already doing, by stimulating creativity by thinking about, for example, how to use algorithms to put more data in their hands when creating and conceptualizing recipes; how to use robotic systems to increase the quality and quantity of what’s possible in the kitchen during a dinner rush; how to offer human and robotic collaboration during plating, to enable previously impossible dish designs. Creativity is not an easy way to get things done – just look at El Bulli, the best restaurant in the world that had to close for months of the year to develop its groundbreaking cuisine – and we think we can play a role in supporting it.
A good example is the recipe creation itself. One challenge we have is figuring out how far we should go in making recommendations in recipe creation. We don’t feel like our role is to tell chefs what to cook or how to cook it. We are not trying to replace their experience or knowledge. We’re trying to create a dialogue so that an AI system can tell a chef, for example: the raspberries you found on the market today, a molecular pairing theory says algae would be fine for them; and based on what has traditionally been associated with raspberries in North America, these spicy and tangy ingredients might go well together as well; in the meantime, here are some ingredients that pair well and are local to you; and these ingredients combine well and are dairy free; these ingredients are low in saltâ¦ etc. What do you think ? What does your experience tell you to do with this information? What ingredients will you select and how will you bring them together?
GN: Have there been any unique challenges (or opportunities) when it comes to teaching an AI to work with flavor and taste, areas that straddle an interesting boundary between science and flavor. ‘art?
Dr. Michael Spranger: One challenge is how subjective and specific the flavor is. Taste an apple and you probably have a different perception of its flavor in your mind. But also, you only have one perception of one apple in 7,500 varieties of apples.
It is difficult for a system to capture an individual’s personal experiences with the “apple”, and the specific flavor data for each of the 7,500 apple varieties (and millions of other ingredients around the world) is not available. currently not kept in the same place. So, given this lack of clarity, how do you make recommendations and suggestions regarding flavor and taste?
It’s a huge challenge, but it’s also what AI is uniquely suited to help: Personalization and accumulation are rich tools to explore.
GN: What were the biggest flavor surprises?
Dr. Michael Spranger: A surprise is how, for gourmet chefs, it’s not just the concept of âit tastes goodâ that counts in flavor. They also care about whether the flavor has a good story. Or if the flavor is new. Or if the flavor matches twelve other flavors they have on the menu.
This presents another challenge for AI, which is figuring out the motivation behind a recipeâ¦ maybe you want to create something that’s never been done before. Maybe you want to create something that has already been done by a particularly famous chef. Maybe you want to create something that has been done for centuries in a particular way by a city of people.
In terms of surprising flavor combinations, there are plenty! A personal favorite was chocolate, junmai sake, and cauliflower.