Editor’s note: Shail Khiyara is the CEO of SWARM Engineering, an AI-powered optimization platform that automates complex operational decisions across supply chain, workforce, production, and logistics.
The views expressed in this guest article are the author’s own and do not necessarily represent those of AgFunderNews.
Don Guinnip wakes before dawn each morning to feed cattle on land his family has farmed since 1837. He is 74. Both hips have been replaced with titanium; he estimates he has two years left.
His children have built careers elsewhere, so there was never a real transition plan. When he stops, the farm doesn’t just change hands, it likely ends.
This isn’t an isolated story, and it represents a deeper shift taking place in the agricultural industry. A recent Wall Street Journal feature on American farm succession put a face to the crisis: Don Guinnip, a fifth-generation farmer with no successor and no plan. But the Journal framed it as an economic problem, which it isn’t; it’s a knowledge-transfer failure playing out across every major agricultural economy in the world.
The part no one planned for
Farmers plan everything, from crop rotations and input timing to equipment cycles, but they don’t plan succession.
The reasons are more structural than personal, and the system makes it hard. Children grow up working the farm but without ownership. Succession conversations get delayed, then avoided. By the time transition becomes urgent, it’s too late.
Underneath all of this sits a deeper issue, that knowledge of how to run the operation lives entirely in one person’s head. Forty years of knowledge—which fields flood in a wet spring, how soil behaves under stress, when to hedge and when to hold, which supplier will actually deliver under pressure—is rarely written down, which means none of it is transferable at scale. When the farmer exits, the land remains. But the intelligence that made it productive disappears.
A global patter
In the United States alone, 315 farms filed for bankruptcy in 2025, up 46% from the prior year. The USDA reports there are now more farmers over 75 than under 35.
Congress approved $31 billion in farm bailouts and disaster relief in 2024. But even with that, corn growers are projected to operate at a loss again in 2026.
The same pattern is playing out globally. In Japan, self-employed farmworkers fell 25% in a single five-year period, the largest decline on record. The average Japanese farmer is now 69, and fewer than 30% of farm operators expect to secure a successor within five years.
Across the EU, the average farmer is 57, with 57.6% of farm managers over 55 and only 12% under 40. In Portugal, half of all farmers are over 65; in Spain, the figure is 41%. The European Commission has responded with a target to double the share of young farmers by 2040 and a proposed $3.5 billion generational renewal package. In Ireland, just 4.3% of farmers are under 35. In China, an aging farm workforce is directly accelerating farmland abandonment.
Where AI actually matters
Agriculture isn’t slow to adopt technology, but it is allergic to nonsense, and that’s a compliment. Farmers reject technology when it has not earned their trust.
Precision agriculture has existed for two decades, with GPS-guided equipment, drone imagery, soil sensors and more where adoption is growing. But precision ag addresses the physical layer of farming and it optimizes execution. It doesn’t help farmers decide when reality stops matching the plan. When the weather shifts, the market moves, or a supplier falls through is where spreadsheets and static forecasts fall apart, and where most operations are still running on experience, instinct, and hope.
What AI actually does when it’s built for this industry, and not just dropped into it, is something different. It starts listening before it decides. It takes the weather data, the price shift, the changed truck schedules, and surfaces the best move in time to act, not after the fact.
Think about what agriculture actually is: biology, logistics, weather, and markets all moving at once. Volatility is not necessarily the core risk to manage but the environment is. And for the first time, we have systems that can reason through that volatility in real time, not replacing the farmer’s judgment but pairing it with speed and computation in a way that was not possible before.
What this looks like in practice
This is where the argument moves from concept to operational reality.
A mill operator no longer relies solely on tribal knowledge to balance throughput and quality. The system has already simulated hundreds of scenarios based on current inputs, constraints, and downstream demand, surfacing the best path before a decision is made.
This means the grain operator doesn’t have to wait for end-of-day reports to react to moisture variability because the system detects it in real time, adjusts routing, and prevents value loss before it compounds into something costly and irreversible.
A new entrant leasing fragmented acreage can therefore test rotations, pricing strategies, and yield scenarios against real data before committing capital and doesn’t need thirty years of experience to make a credible planting decision.
This is what I mean when I say AI stops being a tool and starts becoming a partner, giving the next generation of operators access to the same decision power that previously took a lifetime to build.
The consolidation risk
Without this shift, the default trajectory is consolidation. Guinnip himself predicted it: a future in which farmers work land they don’t own, carry debt they didn’t choose, and lose the pride that came from stewardship.
Large operators and institutional landowners are already moving aggressively into the gap. They have capital, scale, and access to technology that individual family farms cannot match. If AI remains something only the biggest players can access, it will only accelerate consolidation.
The more interesting question, and the one the agrifoodtech sector hasn’t fully answered yet, is whether AI can be the thing that makes the next generation of independent operators viable. We’re talking about people who never had the luxury of growing up on a farm. The agronomist who knows the science but not the generational shortcuts. The entrepreneur who sees the opportunity but not the path in. The investor sitting on land value with no idea how to unlock it operationally.
There are more people who want to farm than the current system can absorb. Remove the knowledge barrier and the equation changes.
What needs to happen
We know the technology exists, and I see the components every day, from agronomic data, and market signals to operational constraints and financial modeling. And while we certainly don’t need another dashboard, what is missing is integration at the decision level: systems that connect all of that into something a working farmer or a new entrant can actually use in real time. Something that moves the needle in yield, cost, time and trust.
The agricultural technology sector has spent decades optimizing yield. The next phase is optimizing decisions across the full operating lifecycle of a farm, including the decisions that determine whether the farm survives a generational transition at all.
Don Guinnip’s farm has been in his family for 188 years. It survived the American Civil War, the Dust Bowl, and decades of commodity price volatility.
It may not survive the loss of the intelligence that sustained it.
AI doesn’t solve agriculture, but it may be the only path to preserving the intelligence that makes agriculture work.
