Once I moved out of my house, most of my dinners were sourced through Skip The Dishes and Uber Eats. The effects of repeatedly eating processed food and the cheap oils used in restaurants didn’t show for two years. Once I moved back home, I decided to get back into shape, so I took some time to read up on various parts of the cooking process and filled my pantry with healthy cooking oils and nutritious ingredients. Being a vegetarian and having eaten my mom’s Indian cooking all my life, I continued the tradition and started making dal and curry at home. Learning my mom’s recipes seemed a daunting task at first, as she would tell me all kinds of tricks to get the most out of the ingredients and rattle off the recipe quickly but putting it into practice was not too bad. The hardest part was cutting the vegetables.
My mom is a great cook. She’s built up a repertoire of tips and tricks for fixing issues and knows the herbs and spices that can work as alternatives to what’s in the recipe. I could’ve found this information online too. Blogposts with recipes will sometimes give helpful information, but the information is scarce and hidden behind advertisements, sponsorships, and lengthy background stories that alter the post just enough to avoid IP issues. After cooking with different methods and making various dishes, I glimpsed that I could modularize each part. When cooking a meal, I found that each ingredient goes through a pipeline and that the combination of ingredients used varies by the method utilized. The important information in the recipe is the events in the pipeline - the action in the current step and the trigger for the next step.
Asynchronously, I’ve been learning about machine learning and web development. I learned of the exciting applications of graph algorithms and the semantic web concept, web 3.0. In the semantic web, all data is machine-readable. All data has some metadata that transfers information about the data, such as the history of its interaction with various objects. To understand how this relates to my cookbook, let’s look at a recipe for making an omelette.
The recipe in the picture above provides step-by-step instructions on making an omelette, with its ingredients, methods, and yield. But since the information was all written out for precisely that recipe, a recommendation system won’t be able to provide alternatives suited to your taste. If you wanted to replace herbs with masala or butter with ghee, you’d have to figure out the differences yourself; there is no semantic information available to suggest alterations.
Instead, if the recipe carries data for each ingredient and all the methods used, a computer would read the metadata and provide personalized recipes for each user. The recipe would be easier to understand and provide nutritional information relevant to the current alteration of the recipe. If we transfer information about the equipment, ingredients, and methods used, we can visualize the events that trigger the next steps.
Additionally, we can alter the view of the graph to various visual representations such as GANTT charts for complex meal planning, pie charts for nutrition planning, or plain text for traditional reading. The below pictures are a rough draft of the idea, visualizing the omelette recipe, plus a bonus recipe of my mom’s gobi-manchurian. Can you see how you would make your alterations and personalize the recipes to fit your kitchen?
A cookbook of personalized recipes leads to a unique digital property that we can use cryptocurrency to manage. A smart contract will store recipe NFTs, with the attributes included in the flowchart, on the blockchain to enable recipe composability as cooks try new ideas and adapt cooking styles. The website’s cooking interface would show the action to take on a card and the trigger for the next step that you can scroll through as you cook.
Additionally, the contract will store a ledger of the transactions between users and their cookbooks to track user interactions and help the best ideas flourish. We can measure the quality of the recipe using interactions such as likes, comments, shares, and ratings while measuring additional values such as transaction volume, momentum, and tips as a gauge for community interest in recipe ideas.
Digital ownership means we can add personality to our recipe collections, work together to compose meals, localize the recipe to personal tastes or local ingredients, and create serendipitous community interactions. You can create your meta kitchen that shows your creativity and share your cooking experiences. And you can view other chefs’ cookbooks to learn from them and compose on top of their ideas. Blockchains use tokens to point the recipes to their creators and the creators to their community to enable these interactions.
On occasions for celebration, we share food and exchange gifts as a token of appreciation. There’s a gift for the person whose occasion we’re celebrating, and there’s a return gift for the guests who came to celebrate. With crypto, we can do the same! If I visit your cookbook and interact with you by leaving a like or a comment or sending a tip as thanks for your insight, you can send a customized digital token as a return gift to show appreciation.
Because the tokens live on the blockchain, we can program the tokens to give access to community interaction and create options for recipe composability. Recipe owners can combine various recipes to create a meal feed and share their meal ideas. Cooks who view these ideas and make their variations on the meal can appreciate these ideas by minting and displaying their gift-tokens alongside their cookbooks.
In creating this application, I am testing two hypotheses. The first is that people will add their unique twists to recipes and that users are inclined to view these unique perspectives and explore the differences to find their preferences. The second is that recipe ownership will enable chefs to work together when creating recipes, meal plans, and cookbooks.
A graph database solution will make the application feature-rich as the information travels with each ingredient and cooking method. Each recipe carries nutritional information and tips and tricks specific to ingredients, methods, and events in the recipe.
Composability will make it easier to build communities as we create, share, or alter recipes when adding them to a personal collection. The application provides digital space, a meta kitchen, for users to showcase individual cookbooks and share their ideas online to collaborate on meals and interact with the community.
The protocol will allow users to fork others’ recipes and make their alterations. The cryptocurrency and digital property ownership advantages that enrich community interactions are: private identity, unbundling curation, digital transactions, and augmented communications.
What excites me about this project is the variety of ways a semantic database of recipes can initiate serendipitous interactions. Provided a marketplace of recipes with nutrition values and preparation instructions, users will create decentralized cookbooks with novel dishes that integrate international styles. The logic of the graph will ensure that all the ingredients get the proper preparation instructions.
Additionally, we can combine diverse applications and products to augment the cooking experience, such as an IoT cookbook for the kitchen that simulates the cooking process as the user cooks. New chefs could learn to cook using AR/VR.
Finally, when we talk about food at my home, we talk about what recipes work well as a combination, how we alter the recipe to make it work with other dishes, or how we made the dish our specialty. I am looking to emulate those conversations with tokens that point to the meals we compose in our meta kitchen with this application.
I hope to bring this app to young adults who want to bring traditional cooking methods into their lifestyle, collaborate with family, and cook delicious meals at home. My vision is to create a community that shares unique meal ideas and simplifies the cooking process. The application creates meta kitchens that capture and recreate the events used to make the recipes in the kitchen.
Thanks for taking the time to read my post. If you would like to sign up for early access, please fill out this form here; I also appreciate any feedback. Follow me on Twitter @duvvuri_krishna to stay updated on my progress with this app. You can also view my progress on my GitHub repository.