A Computer Can Design an Original Recipe That’s Not Totally Disgusting

Both pro and at-home chefs may someday turn to algorithms for inspiration

One of the computer-generated creations, an Ecuadorian strawberry dessert.
One of the computer-generated creations, an Ecuadorian strawberry dessert. Lav Varshney, IBM

A new computer program, designed by IBM, attempts to mimic a creative process usually reserved for professional chefs: crafting unique new recipes by combining surprising or unusual ingredients. The computer algorithms mathematically analyze “vast quantities” of data spanning recipes, cultures and food types, Wired writes.

The program also takes into account the molecular components of different foods in order to avoid chemically incompatible pairings. Wired elaborates on how the IBM team went about doing this: 

The researchers used natural language processing algorithms to scan and parse the text of millions of different recipes. Using this data, they convert a written recipe into a web of relationships, including the quantities of different ingredients and the processes that transform these ingredients into food. They also scanned Wikipedia to learn which ingredients are commonly used in various regional cuisines. They went through handbooks of flavor ingredients to learn which molecules are present in different food ingredients, and also included information about the chemical structures of these molecules. They also included data on how humans rate the ‘pleasantness’ of 70 different chemical compounds.

To use the program, cooks first pick a baseline ingredient, such as pork. From there, they chose a country whose style they would like to mimic. Finally, they select the food genre they have in mind, such as salad, soup or pie. The program spits out a list of potential recipes, which it ranks according to surprise, flavor pairings and “pleasantness of odor,” Wired says.

“We’re pushing the limits of . . . anything that I’ve ever cooked or any possible combination of ingredients that my mind has come up with,” one of the project’s chef collaborators, James Briscione, said in a video interview. At the Institute of Culinary Education, the team has created dishes such as Spanish almond crescents and Ecuadorian strawberry desserts, Wired writes.

The program has application outside of professional kitchens as well. One of the researchers asked his mom to give the program a try for cooking dinner one evening. Selecting her variables, she didn’t make it easy for the computer: “Brussels sprouts” and “Kenya.” The computer spat out a recipe for Kenyan Brussels sprout gratin. She put the recipe to the test, and deemed the results a delicious success. Perhaps at-home chefs bored with the same old spaghetti and meatballs or looking to add a creative flare to their kitchen concoctions could someday turn to algorithms for inspiration. 

More from Smithsonian.com:

 Cooking May Have Driven Human Evolution
Cooking With the Bible

Get the latest stories in your inbox every weekday.