You do not need to remember the workshop. You need one real problem, a bit of patience, and the habit of checking the work. Most of what makes this feel difficult at the beginning is not the technology. It is that nobody tells you what the first month is actually like.
Where to start
Four pages, in this order, if you are starting from nothing.
-
Start here
Get one tool working and build something small in a single sitting, including the first message worth sending to an AI.
-
What to expect
The honest shape of the first weeks: the magic, the confusion, the cost, and the parts nobody demos.
-
Checking the work
How to test something you built when you have never tested anything, and how to tell when an answer is being bluffed.
-
Getting better
A ladder from your first prompt to a tool other people can depend on, with something to try at each level.
The five habits worth keeping
- Start with the problem. Describe the friction in plain language before you name a technology.
- Plan first. Ask for a plan before code, then read the plan.
- Build small. The smallest version that helps one person with one task this week.
- Verify the work. Check the facts, check that it runs, check who it affects.
- Keep a history. Save each working version so your experiments stay reversible.
The reference shelf
For looking things up rather than reading in order.
-
The field guide
The ideas worth keeping from the two-part workshop, in short form.
-
How the pieces fit together
From a real problem to a working system, layer by layer, and which layer to look at when something goes wrong.
-
What you could build
Practical starting points for farms, associations, and rural communities, with a first version for each.
-
A small toolkit worth knowing
A short map of models, agents, and development tools. A map, not a shopping list.
-
Glossary
The vocabulary you are most likely to run into, defined briefly enough to keep moving.
-
Workshop slides
The full deck from both sessions, viewable and downloadable.
Keep the conversation going
The Future Herd Discord is the support layer for the workshop: ask questions, share what you are building, compare tools, troubleshoot problems, and get help deciding what to try next. There is no required pace of participation. Half-finished experiments are welcome, because that is the normal state of a project in progress.
About the workshop
Agrifood Vibe Coding was a two-part workshop hosted on the Future Herd Discord in partnership with the Agricultural Adaptation Council. The first session covered the shape of the AI ecosystem, how models handle language, open source, and how models become agents. The second worked through software development practice, the terminal, development and production environments, and a set of practical things to build.
The slides from both sessions are here. The material that follows keeps the workshop's framing: use AI to extend judgement and capability, while keeping people responsible for goals, evidence, verification, and consequences.
Start with curiosity. Build with care. Keep the history. Share what you learn.