Algorithmic Cooks

Time
2020
Collaborator
Mimi Jiao
Medium
Runway GPT-2, python
Context
Algorithmic Cooks is an exploration into how technology can shape food culture. It is a compilation of machine learning generated recipes about physicalizing our food culture homogenization, and inspecting it with our taste buds. While GPT-2 text generation still feels unrealistic and impractical, it allows us to train a model with a dataset of recipes scraped from different recipe websites to generate innovative recipes based on those we are familiar with.
Algorithmic Cooks is an exploration into how technology can shape food culture. It is a compilation of machine learning generated recipes about physicalizing our food culture homogenization, and inspecting it with our taste buds. While GPT-2 text generation still feels unrealistic and impractical, it allows us to train a model with a dataset of recipes scraped from different recipe websites to generate innovative recipes based on those we are familiar with.


Data Collection
I scraped recipes from the major recipe websites and small regional recipe websites, and then fed them into Runway GPT-2 text generation to train a model to generate recipes. The learning outcome of different unique recipes loses the cultural essence that recipes originally bear with.
Recipes



