top of page
HOT CROPS
Using AI to predict, detect and showcase
the impact of Climate Change on plant crops.
A project developed during oioschool, a short course on AI, Coding, and Interaction Design.
01 The Brief.
Using AI to create a hot object - an interaction, game, tool, or anything else that relates to the theme of heat.
02 The Idea.
The inspiration came from Future Food stories, an exhibition held at Kew Gardens in London.
The exhibition showed how human diets would have changed in the future, introducing new types of food.
It also aimed to highlight the vulnerabilities in the current global food system caused by climate change, pests, and biodiversity loss, while exploring alternative, sustainable, and future-proof food sources for 2050.
The idea behind Hot Crops focused on AI-generated predictions on how plant crops would have to develop and evolve in order to survive the rising global temperatures.
03 The Process.
1. Object Recognition + Definition of 5 scenarios + Evolution Prediction
The hot object needed to recognise each crop using object recognition with Ml5.js, and generate an evolution prediction based on the selected rising temperature scenario with Cohere and ChatGPT .
Scenario Framework
+2.0°
+2.5°
+3.0°
+3.5°
+4.0°
2. Image Generation
Based on the prediction, we gathered an image prompt with Cohere and ChatGPT to use on Midjourney and generate some alien-looking fruits and vegetables. As the scenario increased, the crops started looking progressively eerie.
3. Making it work
We generated the images of 3 fruits and vegetables within every scenario, and proceeded to connect them with the object recognition + evolution prediction mechanism using VSCode.
And voilà, we created a functioning hot object that was able to recognise the plant crop shown on camera and showcase a visual and written prediction of how the plant crop would have had to evolve to survive based on the scenario selected.
We used p5.js to display the prototype and allow users to interact with the object.
bottom of page