Pastry is as much a science as it is an art. Maybe to find out if the first is the most important, Google Cloud AI take a Great British Bake Off winner in a dessert face-to-face. Sara Robinson, hobbyist baker and advocate for Google Cloud developers, created a machine learning model that looked at hundreds of baking recipes (including those for cupcakes, cookies, and scones) to help him find a new one.
The model generated lists of ingredients and quantities that served as the basis for the step-by-step recipes. The model was able to come up with hybrid recipes, and Robinson opted for a recipe that contained a machine-learning-generated cake batter in addition to a machine-learning-generated cookie.
She experimented with patterning and baking to create a desert that incorporates Maltesers, a malt chocolate candy similar to Whoppers. The result was Maltesers AI Cakes (4d6172730a) – the last bit is hexadecimal for “Mars”, which has the Maltesers mark. The company had approached Google to create an AI-based dessert. Meanwhile, Peter Sawkins, the 2020 Great British Bake Off champion, created his own Maltesers cake to compete with the AI’s as part of a promotional campaign.
Robinson and Sawkins both used slightly unusual ingredients to balance the sweetness of the chocolate with salt and umami. Robinson added a pot-infused buttercream frosting after Google Trends identified “Is the pot sweet or salty” as a top search for “sweet and salty.” Sawkins, on the other hand, added soy sauce to a recipe he adapted from one of his mother’s creations. You can Download both recipes and cook them at home to find out which one is your favorite, or have the AI find a recipe for you on this promotional website.
Of course, AI systems are extremely useful for things like wildlife tracking and guide the visually impaired. But, to generate better dessert recipes than those mortals can create on their own, this model will have to learn to omit the Pot. Yuck.