I’m part of a generation caught in between – too young to have grown up fully offline, but old enough to remember life without it. Cellphones were Motorola Razors, no tablets, and the televisions weren’t smart. I bet a lot of you reading this just laughed. Go ahead, but I did get my bachelor’s degree before AI tools were helping with homework and I simply didn’t trust it in grad school, the topics were too specific. And while I have learned to use them some, I also try to do enough research to know when they’re inaccurate.
Still, I see the potential. There’s something to be appreciated about having data-driven tools to assist with making decisions and improving operations. That’s what we’ve been doing all along, isn’t it?
The woman in the post pulled accurate data from reputable sources, including phenotypic and genetic information from bull catalogs and semen sales websites. In theory it makes sense. But, even with that, there’s a limit. That limit is often the lack of human experience.
AI tools can process information, but they don’t understand context like a good cattleperson does. It takes a knowledgeable human to describe those cattle – structure, temperament, how they move, and how they’ve performed over time. Those aren’t things you can always find on a spreadsheet, and certainly not something AI can summarize well without someone knowledgeable guiding it.
It's often stressed that AI is also only as good as the prompts it is given. If you don’t know the information being fed into it, the output won’t be reliable. Asking the right questions also matters – it’s not magic.
That was made even clearer to me at a conference I just attended. One session focused specifically on AI.
One highlight featured an AI-attempted depiction of the daily life of dairy farmers. The results were mixed, especially when it came to the visual representation of the “farm manager’s” office. Let’s just say, I haven’t seen a real life one that would match. But it proved a point – AI still needs some oversight, especially when delving into the realities of some lifestyles.
Where AI can help is in areas like time management and long-term planning. On the farm, the day-to-day is packed. One speaker emphasized that not every task is worth the time it takes, and sometimes AI can help identify where the effort matters most. Another speaker made the point that tractors had auto steer before the cars on the road did, after all.
That said, I’m a little old fashioned. I believe in learning through experiences even if it means making mistakes. If a breeding decision goes sideways, I’d rather it be on me than a flawed AI prompt. That way, I’ve learned something. No one said I – or my bank account – would enjoy it, but it at least means more. There’s value in having the knowledge and researching to gain it if you don’t have it – as well as making the mistakes.
We keep saying we’re still “figuring out” AI, but we’ve been saying that for quite some time. Maybe farming and AI have more in common than we think – both constantly evolve, and neither offer certainty. Time equals trust, so does change.
The goal shouldn’t be to replace instinct or research with automation, but to find a balance.
As for my thoughts, there likely is some potential benefit to using the tools of AI in combination of how we would traditionally make breeding decisions. I try to adapt innovative ideas, though I don’t get to do much of that because of the financial implications in most cases when it comes to farming.
Samantha StammSamantha Stamm is the 2025 Hoard’s Dairyman editorial intern. She co-owns and manages an Angus seedstock and commercial cow-calf operation with her family in northeast Kentucky. Stamm earned a master's degree in agricultural communications from Oklahoma State University and a bachelor's degree in agribusiness with a dual major in animal science from Morehead State University.
