[D] Are LLMs suitable for non-language modeling (e.g. agriculture-climate modeling)?

Hi all,

I am writing a research proposal about foundation models for agricultural modeling. One of the ideas is to use pre-trained foundation models (mainly LLMs like Llama or GPT family). We want to create a farm agent based on climate, crop, soil conditions, and images to make predictions and suggestions for optimal farm management. For the high-tech US farmer, but also a simple farmer in emerging countries/continents (India, Africa, etc.)

Example: Farmer in Kansas asks: What is the best …

agent agriculture climate climate modeling family foundation gpt ideas images language llama llms machinelearning modeling predictions research writing