You probably didn't think much about it when the co-op started using satellite imagery to map your fields. Or when the computer-controlled machine at the fab shop started running programs that used to take three guys to set up. The new automated milking system at your neighbor's dairy just seemed like a smart investment. And when the grocery store put in a self-checkout kiosk, you grumbled the way everyone grumbles about self-checkout kiosks, and moved on with your day.
None of that felt like a revolution. It just felt like the next thing.
A few weeks ago, two of the biggest artificial intelligence companies in the world released new systems on the same day. The new horsepower behind these systems was immediately concerning. A tech founder described asking AI to build a piece of software that would normally take a team of engineers a week. He walked away. Four hours later, it was built, tested, revised, tested again, and finished. Untouched by a human. Done better, he wrote, than he could have done it himself.
Here is something worth knowing. AI can already pass the bar exam. Not barely. It scores in the 90th percentile. It passes the CPA exam. It passes all three parts of the medical licensing boards. Not ten years from now. Right now, today. The AI you can use this afternoon, for free, is the least capable AI you will ever use. It only gets better from here.
Almost every conversation about AI right now is about which jobs it will take. White-collar jobs. Blue-collar jobs. The dispatcher, the estimator, the loan officer, the paralegal, the billing department at the clinic. There are serious people making serious predictions that half of all essential office work could disappear within five years.
Think about what that office work actually means in a town this size. It means the person at the agency who processes your insurance claim. The assistant at the bank who pulls together the paperwork on your operating loan. The bookkeeper who reconciles accounts at the elevator. You know these people. Many of them have spent decades mastering these roles. And there are not a lot of other places in town for them to go. When a bank in Charlotte can process every ag loan in the county with software that never sleeps, the question of what happens to the person who used to do that work here is not theoretical. It is specific, and it is coming.
That deserves to be taken seriously. But it might be the wrong conversation.
When the microscope arrived, the entire field of medicine changed. Problems nobody knew existed became solvable, and the workforce that grew up around solving them dwarfed anything that came before. Nobody peering through that first lens could have imagined the MRI machine that CentraCare installed last spring, or the technicians who run it, or the fact that it uses artificial intelligence to produce images that would have been impossible five years ago. Entire disciplines and entire careers born because a tool made the invisible visible.
It does not think. It does not care. It finds patterns in data at a scale and speed that makes the invisible visible. Think about what that means for problems we currently accept as unsolvable.
We watch a lake like Koronis change over decades and piece together the story from fragments. A water sample here, a fish count there, a guess about what the runoff is doing this season. Now imagine every data point from every lake in the state cross-referenced with land use, soil composition, rainfall, crop cycles, and chemical application records going back decades. AI systems are doing exactly this kind of work right now, predicting algae blooms before they happen, tracing contamination to its source, showing lake associations and watershed districts for the first time what is actually driving the changes they have been watching for years.
Or take the soil. Studies are showing that AI-driven precision agriculture can cut fertilizer use by nearly 30 percent and reduce nitrogen runoff by 35 percent without losing yield. That is money back in a farmer's pocket and less phosphorus in the lake. Same land. Same farmer. Better information.
The answer, if history is any guide, is that there are more of those jobs than anyone currently imagines.
Seeing the problem is the easy part.
Someone still has to look at what the data reveals and decide it matters. Someone has to show up at the township meeting and argue for a different approach. Someone has to look a farmer in the eye and explain what the data means for his operation specifically, in language that respects what he already knows about his own land. And the hard truth is that our local institutions are already stretched thin. City councils and township boards and school districts are running on volunteer hours and tight budgets. A tool that can show us everything wrong with the lake does not help if nobody has the time or the funding to figure out what to do about it.
The opportunity is real. It will not organize itself.
A disclosure. For 25 years, technology consulting has been my career. Helping businesses navigate this kind of change is what I do. These days much of that work involves AI, and I will tell you something that is uncomfortable to say out loud. I do not know exactly what my job looks like in ten years. The tool I help businesses adopt will one day do many of the tasks I currently do, probably faster and certainly cheaper. I have stopped thinking of that as the end of the story. The tasks will change. They always do. The job is understanding what a business or a community actually needs and making sure the tool serves that.
The same is true across this town. Walk into the service bay at the dealership and watch a technician pull a diagnostic readout off a combine. The software flags every sensor reading that looks wrong. The technician is the one who knows that this particular farmer runs his header low because of the grade on his north fields, and that reading is normal. Without that person, the software generates a repair ticket that wastes everyone's time and money. With that person, it catches the one reading that actually matters before a farmer is dead in the field during harvest. The bookkeeper who understands the co-op's business better than any algorithm is the person who makes the software useful. Knowledge of this place is the thing no one in San Francisco can replicate.
But not every job in town works that way. The person whose work is mostly data entry, mostly filing, mostly processing forms that a machine can now process faster and without error. Reassurance does not help that person. A plan does.
That might sound like a contradiction. AI threatens this town and we should use it?
We have been here before. In the 1930s, the power companies looked at rural Minnesota and decided it was not worth running the lines. Too few customers. Too spread out. Not profitable. If farmers had waited for the utilities to show up, they would have been milking by lantern light for another generation. Instead they organized. They formed cooperatives. They ran the lines themselves. And what came through those wires changed everything. Refrigeration. Milking machines. The cream separator that turned a subsistence dairy into a business. The radio that connected an isolated farm to the rest of the world. Entire industries that could not have existed without electricity.
Nobody wiring their barn in 1938 was thinking about the candle business. They were thinking about what they could finally do that they could not do before.
And the question is the same one it was in 1935: Does this community organize and reach for what is possible, or does it wait for someone else to decide whether it is worth running the lines out here? If we wait, we will just be paying a monthly subscription to a company that doesn't know our names.
So what do we do?
Start with the big table. Right now the people writing the rules for AI have never driven a grain truck. Not one of them knows what it means when a creamery closes and the nearest grocery store is thirty miles away. Not one of them is thinking about what happens when a bank branch consolidates because software can do the work. The rules are being written in San Francisco by people who launch rockets into space as a hobby. Washington has barely started paying attention. And there is an empty chair where rural America should be sitting.
School boards should be asking what AI means for the next generation of graduates. City councils should be asking what it means for their tax base. The legislature should be asking who speaks for Greater Minnesota when the guardrails get written.
Closer to home, there are things a community this size can do right now. The co-op can start running AI against its own data to find efficiencies it did not know were there. But there is a catch. If that data is owned by a tech company in California, we have just traded our privacy for a report. True local control means the community and the co-op own the data and the tools. The lake association can push for the kind of AI-assisted monitoring that is already transforming water management in other states. The school can start teaching students how to use these tools and how to think about what they are for. The dealership, the clinic, the bank can all ask: what does this technology make possible that we could not do before?
And then there is the thing each of us can do, starting today. ChatGPT is free. The most powerful general-purpose tool released in a generation costs nothing to sit down with. Most people in this town have never opened it. You do not need a computer science degree. You do not need to understand how it works. You need to type a question. Ask it about your crop rotation. Ask it to explain your kid's math homework. Ask it what your rights are if your bank branch closes. A word of caution. AI is a confident guesser, not an expert. It can be spectacularly wrong about simple facts. Use it the way you would use a sharp new hire who reads fast and talks too much. Check the work. But do not let that stop you from sitting down with it. You cannot have a voice in how this technology gets used if you have never touched it. The people who understand a tool get to shape how it is used. Everyone else gets shaped by it.
Every previous wave of change that hit communities like ours arrived the same way. The technology moved fast. The rules caught up slow. Think about the family farms that consolidated in the 1980s. The main streets that couldn't compete with Walmart in the 1990s. The local newspapers that vanished in the 2000s. Each time, by the time we organized a response, the game was already decided.
You do not need to imagine this. You watched it happen on Highway 23.
When MnDOT drew up plans for the bypass, there were a number of different routes on the table. One of them would have taken the highway from Roscoe straight west to Hawick, eight or ten miles from town. If Paynesville had sat that one out, the road would have missed us entirely. Instead, the city council fought. Business owners showed up. Engineers heard from people who actually live here. The community got five exits and a bypass that touches the edge of town.
That mattered. It was the difference between a setback and a catastrophe. But downtown still lost. When ten thousand cars a day went through town, 1 percent of them stopped. When the traffic moved to the bypass, 1 percent of almost nothing is almost nothing. The storefronts became insurance offices and chiropractors. The retail walked out to Veterans Drive.
Paynesville engaged with the bypass, and it still cost us. Imagine if we had not shown up at all.
We can do better this time. A genuinely new kind of tool has arrived. Someone is going to decide what it gets used for.
There is a lot more at stake than irritating AI-generated content on social media and students cheating on their essay assignments. Our local economy is going to be rewired. We can wait for that to happen to us. Or we can walk out and meet it.
