The “Age of AI” has already begun, and AI has entered our lives without us even knowing it. So to think about what we need and what we can do to design a better future, let’s first look at what will happen to our society as AI becomes more prevalent.
The “future” is already here
In 2015, a company that develops robot kitchens made a splash when it released a video online demonstrating its developing system. It is like a silent drama.
A middle-aged man chooses “Grandma’s Spaghetti Bolognese” from the menu on the monitor on the wall of what appears to be his kitchen. Then, a robot with only two arms and no head or torso moved back and forth on the cooking table like a human arm and started cooking. While boiling spaghetti in a pot, it chopped onions and fried them in a pan with ground meat to make the meat sauce and sprinkled parsley to finish. Place the boiled spaghetti on a plate, pour the meat sauce over it, and voila! The man takes it to the middle of the family table, where there is a lot of chatter, and everyone says, “Let’s eat! “
The scene changed to the chef cooking. There, sensors read the movements of the chef’s arms in real-time, recognizing the angle of his arms, fingertips, and what he is holding, and recording the details. When the chef finished cooking, he clicked the upload button on the monitor, saying, “it’s done Asian salad.”
The scene then switches to a young man’s home. He is relaxing on his sofa, looking at his tablet, when he receives a notification that a new recipe has been released. So he taps on “Asian Salad” from the menu and sends it to the woman who seems to be his girlfriend. The two-armed robot in the woman’s kitchen starts cooking, and just as it finishes, the man comes in after work with a bouquet. And they both say, “let’s eat!” The two-armed robot even cleans up the kitchen counter, which is a bonus.
This video shows that this kitchen robot can cut, mix, fry, boil, serve, and clean up. The system also indicates that recipes can be added from contracted chefs worldwide, and users can search for recipes and schedule cooking according to the time.
The owner of the company that made this video said that their goal was to launch the product in 2019 and that they wanted to be able to offer this robotic kitchen system at the same price point as the system kitchen in a standard home. It was a little delayed from there, but in December 2020, they finally announced that they had come to sell it. Unfortunately, the price seemed to be higher than we had expected.
Still more to come
“Research and development of new technologies related to food, such as food printers and cultured meat, is progressing.
A 3D printer is a device that forms three-dimensional objects in the same way that ink to print letters and pictures on paper. At present, inexpensive and easy-to-use 3D printers for home use are available. On the other hand, food printers use “edible inks” such as chocolate and sauces instead of resins, plaster, and create three-dimensional dishes and add color to them.
One research and development team’s goal is to “print” a hamburger with the ease of a microwave oven, which is currently in many homes. The food printer would not only be able to adapt to the needs of users with allergies or other dietary restrictions, but it would also eliminate the problem of food waste since it would “print” only what is needed.
There are still many difficulties to overcome in terms of taste, texture, and cooking methods, but I feel that the day we can provide drinks like milk is not far off. For example, we can measure the user’s health status for the day and mix fat, sugar, and vitamins according to the biometric information. Then, if we can make it taste the way we want, it will be good for our health and delicious.
Research and development of cultured meat are progressing rapidly. Cultured meat obtains meat not from individual cows, chickens, fish, etc., but by “culturing the cellular tissue” of edible parts. If this research and development go well, there will be no need to kill animals. In addition, the carbon dioxide and manure produced in growing them will be eliminated, which is good for the environment.
Cow burps and farts accelerate global warming because they emit methane gas. The amount of methane gas emitted by a cow daily is 300 to 500 liters (*Reference). There are 1,000,000 cows in the world, and considering that it takes a long time to raise a large animal, it is understandable that there are high hopes for this research and development. But, of course, we all know that this is not the only cause of global warming.
Meat cultured in a cleanroom is said to be hygienic and safe. It could be a solution to the world’s food shortage problem. In December 2020, newspapers reported that the Singapore government became the first in the world to approve the sale of cultured chicken meat. It is only a matter of time before we can eat chicken nuggets made from cultured chicken meat.
Beating the world chess champion
In applying AI in cooking, a system has already been developed that can suggest dishes that humans have never thought of before, according to the user’s preferences.
In 1997, the world’s largest computer manufacturer developed an AI called Deep Blue that defeated the human world chess champion, Garry Kasparov. Later, this manufacturer created Watson, an AI that understands and learns natural language (words spoken by humans) and helps humans make decisions. And in 2011, Watson won the popular American quiz show “Jeopardy!” 2011.
There is a big difference in technical difficulty between winning at chess and winning at quiz shows. In chess, there is only one opponent. There is only one opponent in chess, and the game’s rules determine the range of moves that can be made. So the number of combinations is enormous but finite and can be calculated. This kind of problem is called a “well-defined problem” in the field of AI. It is a problem of finding an appropriate solution under certain definite conditions and circumstances. The same applies to Othello and Shogi.
On the other hand, questions asked in quiz shows are not always the same in the way they are expressed or the text of the questions. As you can see from watching a battle between humans, the competition is for speed, and if you don’t listen to the question until the end, you may get the meaning of the question wrong. In this situation, memorizing the question-answer pairs will not help you win.
In a quiz show, the first step is to understand the question, and sometimes, in the middle of a question, to predict the content of the following question, or to draw out from a vast amount of “memory” (data) things that may be relevant, or to “create” new knowledge by combining knowledge you already have and deriving it from there. In a sense, creating new knowledge can be a form of creative intelligence, and winning the quiz king made the news meant that AI research had moved to the next stage.
However, there are still a lot of problems we encounter in our daily lives that cannot be solved with this level of intelligence. For example, in research, product development, etc., we are bombarded with problems that do not contain all the information needed to solve them. As a result, no single correct solution can be determined. These are called “ill-defined problems” as opposed to “well-defined problems.” In the future, more research and development will be done on the latter.
AI’s challenge to cooking
A foray followed Watson’s victory over the quiz king into the field of cooking. It analyzed psychological data on flavors, combinations, and preferences (tastes) of ingredients from cuisines worldwide and created new and unique recipes from them. This system is called Chef Watson, and it not only searches for and provides recipes for existing dishes but also creates new dishes that are unique to each regional cuisine, such as “Japanese,” “Italian,” and “Chinese,” and presents them as recipes. In 2015, some of these recipes were selected and published as a collection of recipes created by the AI, along with photos of them prepared by a human chef.
This was not a chess game with a single opponent, nor a quiz with a single answer, but a creative challenge in cooking. In developing this system, the team first defined creativity in cooking as “quality” and “novelty.” Quality” refers to deliciousness, while “novelty” relates to surprises that have never been tasted before, and our goal was to propose delicious recipes with combinations of ingredients and textures that humans have never imagined.
With the resulting system, the user first selects the ingredients they want to use in the dish and then chooses the style they like, such as Chinese, Italian, French, or Vietnamese. Then, by selecting the eating occasion, such as a celebration or an everyday dinner, and inputting other information such as ingredient restrictions and preferences, the system will suggest recipes based on a wealth of data. The entire course menu includes aperitif, appetizer, main dish, and dessert. However, the system only provides the recipes, and it is the human being which makes them. People who have cooked and eaten the recipes can evaluate them and give feedback to the system.
The system uses the ability to interpret human speech (natural language) to capture human experience and knowledge as information. The system then translates it into a language the system understands (programs, algorithms, data, etc.). Finally, it analyzes it to develop a hypothesis that combining this might be tasty, which the system itself evaluates and learns from.
In general, we tend to think that machines do not have creativity and that creativity is a strength that only humans possess. However, if we think of creativity as the ability to propose new combinations of ingredients and cooking methods, we can say creativity here. This system may be more creative because it can offer new possibilities. After all, there are no assumptions or stereotypes about dishes or ingredients.
Technology for solving social issues
I have talked about robots and AI, and we have introduced new technologies related to food. These are undoubtedly dreamy, interesting, and delicious stories to imagine the future! But let’s stop and think for a moment. Can’t we use these technologies not only for dreaming but also for solving problems that are currently happening in the world, in Japan, and in our daily lives?
( to be continued )