Why Your AI Doesn't Know Your Farm (And How to Fix That)

Hey Reader,

Week before last, I talked about how I use AI on the farm and in my business.

This week I want to show you something most people skip entirely - and it's the thing that makes AI actually useful instead of just… meh.

Here's the problem with how most of us use AI.

We open it up, ask a question, get an answer, close the tab.

Next time we come back, it has no idea who we are.

It doesn't know what we raise, how we price things, what our values are, or how our farm operates.

So every answer is generic. And generic answers are only so helpful.

There's a better way.

ChatGPT and Claude both have something called Projects.

Quick note before you dive in - both ChatGPT and Claude offer this feature for free. You don't need to pay anything to set up a project and start building your farm blueprint.

Free gets you started. If you find yourself using it heavily and hitting limits, both paid plans run $20/month. But start free, see if it's useful, then decide (I have no affiliate relationship with either).

Now, on to Projects. Think of it like a folder that stays open.

Everything you put in there, the AI remembers. Every conversation you have inside that project builds on what it already knows about you.

Here's how I'd set it up for your farm.

Step 1: Create a project.

In ChatGPT, click on "Projects" in the left sidebar and hit New Project. In Claude, you'll find Projects in the same area. Name it something simple - like your farm name - works fine.

Step 2: Write a farm blueprint document.

This is the most important step. Before you ever ask it a question, give it a foundation to work from.

Open a simple document and write out (Google docs is free and what I use):

• How you got started and why
• What you raise and sell
• How you raise them (pastured, grain-finished, heritage breeds, etc.)
• Who your customer is
• Your values as a farm
• How you like to communicate (casual, professional, no hype, etc.)
• Any non-negotiables (things you never want it to say or suggest)

It doesn't have to be long. A page is plenty to start. Upload that into your project.

Step 3: Add your real numbers over time.

This is where it gets genuinely useful.

As you go, drop in your price list. Your product descriptions. Your packaging notes. Your CSA details. A copy of your customer welcome email.

Add in philosophical views you have and goals for your farm.

Every time you add something real, the AI gets a little more dialed in to your actual operation.

Step 4: Asking questions and get much better answers.

Now instead of getting generic answers, you get answers that sound like they came from someone who actually knows your farm.

Ask it to write a product description and it'll use your voice, your breed, your farming philosophy. Ask it to help you price a new cut and it'll reference what you've already told it. Ask it to draft a customer email and it won't start from scratch - it'll start from you.

That's the shift. You stop using AI like a search engine and start using it like an advisor who's been paying attention.

It won't replace your judgment and shouldn't. But it stops wasting your time re-explaining yourself every single time and getting generic answers that aren't helpful.

Start with the blueprint. Even a rough one. You can always add to it. Then you're well on your way to having an idea machine tuned to your farm.

Jason

BTW: If you're still taking orders through DMs and texts, AI can't fix that. But we can. Homegrown Hosting builds your farm's online store, writes your copy, enters every product, and keeps everything running — so you can stop managing orders and get back to farming. Learn more here.