Land O’Lakes CTO to Demystify AI at Tech Hub LIVE Keynote

Artificial intelligence generates a lot of excitement across agriculture — and with that excitement comes a lot of questions. Teddy Bekele, SVP & Chief Technology Officer for Land O’Lakes, wants to use his Tech Hub LIVE keynote to pull back the curtain on how the technology actually works, before diving into where it’s already delivering value on the ground.

“Part of what I’m going to try to do in the talk is to demystify AI a little bit, so it doesn’t become this black box,” he says. “Once you understand how it works and what it’s capable or not capable of doing, then we can start to apply how it’s best used today.”

For Bekele, that starts small — using AI to respond to emails, prepare for a meeting, or make sense of data that just came in from the field.

Bekele will deliver the Welcome and Opening Keynote, “Real Impacts of Artificial Intelligence on Food and Agriculture,” at the Tech Hub LIVE Conference, July 20-22, 2026 in Des Moines, Iowa. His session runs Monday, July 20, from 4:30 p.m. to 5:15 p.m. on the Tech Hub LIVE main stage, immediately following the Women in Ag Tech event and ahead of the welcome party. There’s still time to register at https://techhublive.com/register/

Clearing Up the Big Misconceptions

Bekele says there are two misconceptions he’s hoping to clear up during his keynote. The first centers on data quality.

“Having the data is one thing. Having good data is a necessity,” he says. “That’s one thing that, you know, we have the data, and let’s just AI run through it — that’s not that simple. We’ve learned that along the way.”

The second misconception he wants to dispel is the idea that AI will eliminate jobs across agriculture.

“I think AI is going to transform jobs, and new jobs get created, and jobs that are existing today may look different,” he says. “I don’t quite see this idea that the workforce is going to disappear and these things run it.”

That transformation, he adds, requires an active effort. Using AI to augment personal productivity is straightforward, but deploying AI agents that make decisions and perform tasks means rethinking entire workflows.

Real Examples From Land O’Lakes

Rather than lean on hypotheticals, Bekele plans to draw his keynote examples directly from Land O’Lakes’ own AI deployments — including the successes and the missteps along the way.

“I actually want to get through the how did we do this, and why did we get here,” he says. “The ones I really know are the ones we’re doing, because I also know the good and the bad.”

One example is Oz, Land O’Lakes’ small language model, which turned a dense crop protection reference guide into a conversational chatbot.

“The idea is to bring the book to life,” he says. “You can give it the specifics of where you are and what you’re doing, and it’ll give you the proper recommendation. The expectation is not that the agronomist doesn’t know that, but it maybe reaffirms a hunch they had, or they can go a little bit deeper.”

On the feed side, Bekele points to how Land O’Lakes has automated much of its animal-owner order-taking process with AI, while routing complex questions to customer service, then feeding those insights back into the AI so similar issues get resolved before they happen again.

Unlocking Insight That Was Already There

Bekele believes one of AI’s biggest opportunities in agriculture is not new data collection, but finally making full use of the data organizations already have.

He points to years of Land O’Lakes research on magnesium levels in chicken feed and its effect on eggshell strength and salmonella risk — insight that rarely made it past a short label claim.

“We never really explained why that’s important, and why this feed is better than this other one in specific situations,” he says. “I feel like with AI now, we can unlock things like that.”

The same holds true in the field, he says, where yield, as-applied and as-planted data have long existed but haven’t always translated into deeper insight about what to do the next season differently.

Why Trust Takes Time

Balancing innovation with the practical realities of agriculture — farmer needs, trust and adoption — remains central to how Bekele’s team approaches AI deployment.

“The number one piece is trust,” he says. “You have to trust what you’re doing with AI. And then whoever you’re giving it to also has to trust the AI and trust you as well. So it takes a long time to do that.”

Bekele says one model his team built came together quickly, but took far longer to actually deploy.

“We wanted to make sure we had it right, because one wrong answer could have catastrophic effects,” he says. “But two, it also kind of just erodes that trust big time.”

That’s why his team keeps a mixed portfolio — pairing quick, low-risk wins with longer, more experimental projects — so the group builds momentum while still tackling bigger, unproven ideas.

Where to Start

For companies and individuals unsure of where to begin, Bekele encourages experimentation over an all-or-nothing approach.

“The art of prompting and getting the AI to give you the answer back is in itself an art that will get better over time,” he says.

He also suggests starting wherever teams already feel overwhelmed.

“Go to places where people feel overwhelmed with a lot of tasks,” he says. “Are there things they’re doing that are manual in nature that could be alleviated by the use of AI?”

Overcoming the Three I’s

Bekele plans to close his keynote with what he calls the three I’s standing in the way of AI adoption: ignorance, inertia and imagination.

“Ignorance — not in a bad way, but ‘I don’t know how to do this,’” he says. “Just start simple, and go from there, and you’ll see how excited you get, and what more you could do.”

Inertia, he says, is the discipline of actually following through. Imagination is what comes after the momentum builds.

“It’s so awesome when you see people unlock some of the things they think about that are unbelievable,” he says.

“I’m excited to see that sort of flourish, and it’s going to be down to the individual and each organization and how they want to deploy and use AI.”

His overall message to attendees: don’t be afraid of it.

“Learn it, take your time to understand it, make sure you get comfortable with it,” he says. “But at the same time, go work on the things that have the highest value.”