IBM's Watson is a cognitive computer designed to handle complex problems and learn from and interact with humans. And, with the Chef Watson app, it can create novel recipes through an understanding of food science and people's taste preferences.
Chef Watson takes Bon Appetit's database of 9000 recipes and can generate millions of new recipes based on qunitillions of possibilities. The video above explains how it works, but this is why Chef Watson is unique, in a nutshell:
In contrast to a search engine that simply sifts through existing data to serve up a list of already published recipes, Chef Watson is not programmed to come up with a defined answer to a defined question -- it understands, learns, and considers not just the data behind recipes and flavour compounds, but also human perception to design highly creative recipe ideas.
Using the app, users can create novel recipes by selecting ingredients, dish type and dish styles, as well as any ingredients to exclude in the final recipe. Once the user has entered in their choices, the app will generate a list of never-before-seen recipe suggestions.
Recipes created and suggested so far include Russian Celery Sandwich, Salmon Taco, Scottish Gumbo. You can adjust how classic or novel you'd like the recipe to be, save recipes, and share them. The more ingredients and cooking styles you add, the more suggestions Watson will offer.
The app is currently in beta, but you can request access (I got in the same day). There's also a cookbook created by IBM Watson and the Institute of Culinary Education (ICE) if you want even more computer-created recipes to try.
Seems like a fun and geeky way to explore new recipes.