IBM’s AI computer has come up with some pretty incredible food pairings

There is a theory that monkeys can finally output all of Shakespeare's work in unlimited time. But can monkeys come up with a new way to combine saffron and fennel? IBM's Watson Artificial Intelligence Computer is known for winning in Jeopardy. - I'm about to do this

Steve Abrams, the research director of IBM Watson Life, told Quartz that Watz scans public data sources and created recipes, compounds in food, and a vast library of common pairs. Abrams announced that wiki is a very useful resource. Now you can instantly access Michelin star chef with tablet for lifelong knowledge.

Watson differs from a normal computer (or chef) in that it can find patterns from a large amount of data. Through pure repetition, you can basically understand how the combination of compound and food functions synergistically. "Strawberries and mushrooms share many flavor compounds," Abrams said. "I understood that they were doing well."

The Watson team actually cooperated with the College of Culinary Education (ICE) to publish an artificial smart dish recipe and the laboratory was established on April 14. Quartz was unable to test Watson's esoteric game, but it is somewhat prominent:

Watson will not be able to replace any MasterChef participants in the near future - it can only provide ingredients, but the chefs still need to turn them into edible - it casts doubts about what constitutes creativity I will. Abrams called Watson "a cognitive computer" designed to 'stimulate human creativity'. The Watson chef is one step higher than the average recipe suggestion site, but artificial intelligence has not reached the point that it can move on to taking action from computer providing information.

According to Mr. Abrams, Mr. Watson says "It can show cold facts and there are no prejudices," it is different from ordinary people. However, until the robot stands up and accepts our work, humans will make final decisions on food preparation. This may be a good thing.

Please see a more concrete example. Let's think about IBMChef Watson. This cooking version of IBM AI software starts with ingredients and cooking styles and generates a lot of ingredients and recipes. Results are obtained by manipulating and processing information from existing formulations and scientific data on flavor compounds. Interestingly, the Watson chef does not have cultural prejudice and is not afraid to mix food from various sources and mix with unexpected ingredients (with tofu and mozzarella cheese using services Small play - this will definitely irritate my Italian parents And Japanese friends are shaking These unexpected combinations can be a new starting point for creating new recipes .

AI Design Recipe: IBM Watson uses machine learning to make new recipes and cocktails that human chefs can not think of. This algorithm matches food based on a wide range of taste characteristics and the chemical composition of the food in its database. We tried some Watson cocktails last year at A360: Hoochon - Honey ale (sliced ​​beef slices, roast beef and IPA), plum pansetetta cider (bacon and cider) and co - op cope (old chicken). Fuel not happy food: For people who choose to optimize consumption (usually at the expense of "taste"), we need nutrients, carbohydrates, proteins, and our bodies We use a mixture to provide things. . Companies such as Soylent and Ambronite are already doing so. These choices will be cheaper and faster than traditional consumer practices - and they will provide food to many people who simply can not get food.