This work explores a technological answer to the perennial question "What's for dinner?" Few of us know with great certainty the exact food we crave; instead, we stew on the question and explore the nature of cravings with imaginative descriptions: "I feel like something light, fresh, sophisticated, not too mushy; something influenced by Thai or Indian, something aromatic." Synaesthetic Recipes is a visual search program that allows imaginative textual descriptions, using them to drive recipe recommendations. A database of 100,000 recipes is automatically annotated with common sense about food. An artificial intelligence robotic reader reads each recipe, and based on flavors of the ingredients and cooking procedures predicts how a food will look, taste, and smell. We are translating recipes into the rich descriptive vernacular of how people naturally conceptualize their cravings for food.