The Programmable Farm

With the caveat that our sense of the future is near term, and that geopolitics is a separate clusterfuck that makes long term planning difficult, here's our current take.

A few things are converging in agriculture that we think will become genuinely significant over the next few years.

The most intriguing may be vibe coding for agri-food: farmers, food processors and agricultural organizations using conversational AI to build small pieces of software without waiting for a conventional development team. Think custom herd records, irrigation dashboards, traceability tools, equipment logs or market-intelligence systems built around the peculiarities of one operation. Agriculture has always suffered from generic software designed far from the farm. Vibe coding could reverse that relationship by allowing the people closest to the problem to prototype their own tools.

At the same time, drones are moving from novelty to everyday agricultural infrastructure. Their role is expanding from crop imagery into scouting, mapping, spraying, livestock searches and storm or drainage assessment. The interesting development is the service model: many farms may never own a sophisticated drone, just as they do not own every piece of harvesting equipment. Local operators, co-ops and crop advisers can provide intelligence or application by the acre. DJI's latest agriculture report gives a sense of the global scale.

Livestock may be where AI becomes most tangible. Cameras, microphones, wearables and machine learning can increasingly recognize individual animals and detect changes in feeding, movement, rumination, heat, lameness or social behaviour. The important shift is from occasionally inspecting the herd to continuously interpreting it. Recent research includes AI systems for cattle detection in complex farm environments and multimodal models that classify specific cattle behaviours. The practical promise is earlier intervention; the harder questions concern false alarms, farmer attention, animal welfare and who owns the resulting data.

Wireless grazing technology takes this a step further because the system can act as well as observe. GPS collars allow a farmer to draw and move grazing boundaries digitally, while audio and vibration cues guide cattle. Halter has begun satellite-connected virtual fencing in Canada, with its first Alberta ranches now live and broader availability still limited, while university researchers are studying the welfare and productivity implications. This could make rotational grazing, conservation grazing and the management of difficult terrain considerably more practical. It also turns fencing into software — which means reliability, subscriptions and technological dependence become part of grazing management.

Water is likely to tie many of these systems together. Smart water management combines soil-moisture sensing, weather forecasts, flow meters, automated valves and leak detection to apply water when and where it is needed. As drought, extreme rainfall and pumping costs become more consequential, water intelligence may matter as much as yield intelligence. Canadian projects are already using precision systems to reduce water consumption while maintaining crop health. The next frontier is not merely automated irrigation, but systems that understand the whole farm's water movement.

Then there is upcycling, which brings the same logic into food manufacturing. Food companies are beginning to treat pomace, whey, spent grain, imperfect produce and processing remnants as feedstocks rather than waste. The Upcycled Food Association has helped turn this into a recognizable product category and has established a Canadian network. This is interesting because it joins sustainability to product development: the waste stream of one business becomes the ingredient stream of another.

Finally, we would watch the right-to-repair movement. Modern agricultural equipment is increasingly governed by software, diagnostic access and digital locks. While there is legislative movement in this area, the growing accessibility of AI means that people can now hack their technology directly to repair on their own terms.

What connects all of these trends is that the farm is becoming programmable. Drones provide the eyes, sensors and AI provide interpretation, collars and valves provide action, vibe coding lets people design new controls, and upcycling extends the system across the food chain.

That makes the deeper hot topic agricultural agency: who can understand, modify, repair and benefit from these increasingly intelligent systems? The winners may not simply be the farms with the most technology. They may be the farms, organizations and regions with the greatest capacity to adapt technology to local knowledge — and to retain meaningful control over it.