Cargill leans on regenerative agriculture and generative AI to feed the planet

It turned out to be a prognostic moment in Hartsock’s life. Since 2022, she has served as chief information and digital officer at agricultural giant Cargill.

“I wish I could say that at that age, I was like, ‘Someday I will get a chance to go work for Cargill, and then I will be able to marry my passion for technology with helping improve the lives of our farmers,’” jokes Hartsock.

But when an executive recruiter at Cargill called, her Midwestern childhood combined with decades of work in technology compelled Hartsock to answer.

For much of her career, Hartsock worked in the construction equipment and energy sectors, serving as chief information officer during two big mergers. She worked at equipment maker Cameron when it was bought by Schlumberger and at General Electric’s oil and gas equipment business when it merged with Baker Hughes. Hartsock grew accustomed to working with global manufacturing suppliers and long product lead times.

But Cargill, with annual sales of $177 billion, felt different not just because it produces food. Cargill works closely with farmers, a group Hartsock had never worked with before. As CIO, Hartsock needs to ensure that her company’s tech tools and capabilities are serving this constituency as effectively as they serve company employees.

“If the farmers can’t win in whatever we want to deploy…then we feel like we’re not creating sustainable change,” she says.

Data is an important part of this mission. Cargill uses crop data and soil testing to bolster regenerative agriculture practices, which can help lift the profitability of a farm. Cargill’s satellite imagery systems provide alerts on deforestation issues. And a digital tool it developed called CocoaWise provides transparency of the cocoa bean supply chain to help customers determine if the chocolate they are buying aligns with their ESG goals.

“We have commitments that we make with our peers, around traceability and visibility of our data, whether it be sustainable or transparency in the supply chain itself,” says Hartsock.

Investments in artificial intelligence have changed how Cargill forecasts movements within the company’s supply chain, including timing when a shipment across the ocean will reach ports.

When it comes to generative AI, Hartsock sees three focus areas that will create meaningful value: research and development, knowledge management, and productivity. 

For R&D, generative AI could help propel innovation, including ingredient formulation. For knowledge sharing, one early generative AI production case that Cargill is piloting is a chatbot called “CASC Me Anything,” which stands for “Cargill Agriculture Supply Chain.” “If only Cargill knew what Cargill already knows,” says Hartsock of the power of tapping into the company’s historical data. 

Cargill is also experimenting with generative AI to bolster internal productivity—anything from generating computer code to creating the text for job postings—though it’s still early days. The large language models that Cargill is testing include those offered from existing vendor relationships with cloud compute and data storage partners, as well as others outside that existing ecosystem.

“It’s going to be a catalog of partners, some of which we’ll experiment with and love, and some we will experiment with, learn something, and move on,” says Hartsock.

John Kell

Send thoughts or suggestions to CIO Intelligence here.

NEWS PACKETS

Are PCs & AI a perfect match? Not quite yet. Chip manufacturers and personal computer makers have been announcing new products over the last year that allow for AI to run on a PC rather than the cloud. The pitch, the Wall Street Journal reports, is that doing so will lower cloud bills, offer greater security, and aid CIOs as they deploy more AI. But CIOs aren’t yet convinced. “I’d say it’s rather costly,” says Elizabeth Hackenson, CIO of energy management firm Schneider Electric. “Waiting, even a bit, will give us more mature options,” says Jay Ferro, chief information, technology and product officer at clinical research data-management company Clario.

OpenAI & Meta’s AI battle heats up. Mere days after social media giant Meta unveiled updated versions of its free Llama AI model, OpenAI has introduced new enterprise-grade features for building and programming atop its GPT-4 Turbo LLM and other models. VentureBeat reports that OpenAI’s upgrades—with a focus on security, administration control, and cost management—are an indication that OpenAI is more focused on a smoother “plug and play” experience for businesses, while open models tend to require more setup from enterprises.

Disney CTO exits to join Penn Entertainment. The chief technology officer of Walt Disney Co’s entertainment and sports networks has left for the operator of ESPN Bet sportsbook Penn Entertainment, Bloomberg reports. Penn, which has paid Disney $1.5 billion over 10 years for ESPN Bet, is aiming at rivals FanDuel and DraftKings for a bigger piece of the sportsbook market. Penn also operates 43 casinos and horse tracks, as well as online betting in 19 jurisdictions.

ADOPTION CURVE

While it is no surprise that the technology sector is leading in AI projects that are either up and running or in pilot testing (70% of companies), banking & financial services (55%) and manufacturing (50%) aren’t too far behind, a survey by cloud-computing company NetApp showed. Health care (38%) and media and entertainment (25%) were the laggards.

“We saw that 40% of media and entertainment companies and 30% of healthcare companies reported having no AI projects started,” says Jonsi Stefansson, senior vice president and CTO at NetApp. He noted that ethical issues around transparency with viewing audiences, as well as worries about the quality and authenticity of AI-generated content were barriers for the entertainment industry, while health care firms face challenges due to strict regulations that guard patient privacy data.