Everyone is asking the wrong question about AI.
The conversation around AI has been dominated by what it can replace. Jobs, workflows, entire departments.
But that is the wrong frame.
The use cases getting the most attention are the ones you can see and feel: a chatbot that answers questions, a tool that writes your emails, a system that reads documents and tells you what matters.
But the companies that are going to look fundamentally different five years from now are not the ones that figured out how to cut headcount. They are the ones that figured out how to build things they never could before.
Earlier this year, we stood up Side by Side, a new offering of our Crop Plan Warranty. The software behind it was built from scratch. We shipped in under six months. The last project of this scope took us years.
That is not a small difference.
It changes what we say yes to. It changes which ideas ever see the light of day.
How it happened
There is no single thing that explains the speedup. It is a lot of small compressions adding up.
Engineers use AI throughout the build, not just when writing code. It accelerates data exploration and compresses the time spent on internal tooling. Documentation and testing that used to consume weeks barely register now. None of those are dramatic in isolation. But when you spend less time on the bounded, repeatable work and about the same amount of time on the parts that actually require judgment, the total shrinks considerably. You finish faster because less of the build is scaffolding.
The hard parts stay hard. Domain modeling, risk math, the decisions that require genuine expertise in how agriculture actually works. AI does not help much there, and we are not asking it to. What it does is clear the path so that the people doing the hard work are not constantly slowed down by the work around the work.
What it means for what gets built
Every business is constantly deciding what is worth pursuing. Time, capital, and people are limited, so every new initiative comes down to the same question: will the return justify the investment?
For companies that build software, that decision has historically been shaped by engineering cost.
If something is going to take two years to build, it needs to be close to a sure thing before anyone signs off. A lot of good ideas never made the cut, even when the market was plausible and the domain expertise was there.
The investment was too steep to justify without more certainty than anyone could reasonably have before starting.
AI changes that calculation.
Not by replacing engineers, but by making them dramatically more productive on the work that used to fill out most of a build. That changes what gets said yes to. An idea that wouldn’t have survived a two-year commitment can survive a six-month one. More ideas get tried. The cost of being wrong about any one of them is smaller than it used to be. And the ones that do not work get killed before they have cost much.
This is the real shift. AI is not just changing how we build. It is changing what gets built in the first place.
Side by Side is the clearest example we have right now. If it had come with another multi-year commitment attached, it probably would have stayed on the list. At six months, it was an obvious yes. We have spent years building domain knowledge and actuarial and agronomic data. The constraint was never the ideas. It was the bandwidth to build them. That constraint just got a lot looser.
What the ag AI conversation is missing
Most of the AI coverage in agriculture right now is focused on the same layer: grower-facing tools. Variable-rate prescriptions, disease detection from field imagery, AI-powered agronomic recommendations, and precision application systems. These tools are real, the use cases are proven, and the investment is following.
What gets less attention is the shift happening inside companies rather than in front of growers.
The question of which organizations can turn cheaper engineering into actual new businesses is going to matter a great deal over the next five years. That is not a product feature that shows up on a homepage. It is a decision that happens in a planning meeting, when someone asks whether an idea is worth pursuing and the answer changes because the cost of finding out just dropped significantly.
For the companies building tools and programs that serve agriculture, that shift has real implications.
The advantage is not just in having better ideas. It is in having the ability to build them faster than anyone else.
That gap is only going to widen.