A reaction to "What role does AI play in Agriculture?" post by Todd Orman

When I studied AI in computer science 40 years ago the key challenge we explored was the "body problem". Grossly simplified, the concept is that only a minuscule portion of the information present in physical reality can be tokenized in ways that enable it to be represented digitally, let alone processed reliably.

This problem remains, and is hugely apparent in the domain of agriculture.

There are huge unrealized opportunities in application of digital solutions and Machine Learning (ML) to agriculture. But the ChatGPT findings referred to in the this post by Todd Orman reflect why this potential remains unrealized; the unchallenged assumption that the incentives, strategies, and tools that provide returns to urban business will be perceived as returning, and will return, value to farming and ranching.

The incremental production gains reliably realizable from sophisticated data collection and analytics in ranching and farming have always been, and continue to be, overstated. Harsh reality is the single digit improvements realizable from refined management techniques are almost inevitably lost as noise within the chaos that weather, logistics, and markets impose.

ChatGPT proposed that "Manual" and "Routine" machinery operation will be obsolete by 2030. The longest uninterrupted period of "routine" machinery operation I have ever experienced in 30 years of farming is maybe 30 minutes. Machinery overheats, fails, leaks, plugs, breaks, ingests things they were not meant to ingest, makes funny noises, need to be relocated, fuelled, serviced, provisioned and unloaded. Four hours sitting in the nice clean cool cab listening to tunes is the reward for the hours of dirty and physical effort it requires to secure a machine in the field and operating.

"Mechanics & repairs" become specialties? Not on any farm I have ever been on. It is November 3 and the snow is coming in 6 hours; you duct tape and wire the damn thing back together and get going, not wait half a day for a mechanic to show up.

Farmers and ranchers exist in the physical realm. They work outdoors in a physical environment and they deliver physical products. Their fundamental motivation is first sustainment, and hopefully expansion, of a property legacy for the next generation. Their second motivation is stubborn independence. These fundamental truths continue to be overlooked, and often not respected, by those, mostly urban based, who conceive and deliver the majority of digital agriculture solutions, and is a key reason why these solutions continue to struggle to secure adoption.

Addressing these realities forthright is where Olds College can uniquely deliver value in the domain of digital agriculture. The strength of the institution comes from its existence on a working farm in a rural farming community.

I am going to challenge you Jay Steeves . Do not just disregard what ChatGPT told you; reject it. Instead listen to your students, your instructors, your advisors (the Digital Agriculture IAC has not met in nearly a year) and the farmers that surround you and the campus. Work back from their needs and their realities, which is very different from what ChatGPT told you, and what even most Canadian Ag "experts" will tell you.

Address head on, and provide solutions to, producer data sovereignty, and securing fairness in realization of the returns from off-farm application of producer data.

Work with Alberta AgTech entrepreneurs like Flokk Systems Inc. to help us better understand Canadian producer requirements. And even more importantly, apply the Colleges influence and resources to help us ensure we can access the resources we need to deliver those solutions to Canadian ranchers and farmers. I have placed exactly such an opportunity in front of Todd Ormann ; help him realize it.

Explore solutions that work within the constraints rural Canada faces today; poor and expensive connectivity, absence of support resources, and affordability. Rural Canada needs good enough today, not perfect someday.

Understand, advocate, and begin preparations to address that Canadian agriculture is strategic, not tactical, to the nation. Four of the five nations with largest "water gap", the rate by which a nation is drawing down freshwater reserves faster then they are replenishing, are nuclear powers, and two of them are superpowers. Ensuring Canada can feed these nations is not only how Canada will prosper, it is vital if we are to retain our sovereignty.

The ChatGPT list is far too passive; it identifies roles that implement, sustain, and apply tools others, which as Canadians typically means Americans, created. This is no longer sufficient; Canadian success requires we build and sustain our own solutions for our own needs.

Lead the effort to secure Canadian digital sovereignty in Canadian digital agriculture. Open source offers huge, and affordable, opportunities to realize this.

The potential, and success, of OC will not come from listening, and responding, to ChatGPT. It will come from the Colleges gift of being able to directly listen to and engage the ranchers and farmers in your classrooms and your community, and then leading in the creation of, and ensuring the necessary conditions are in place to deliver and sustain, the solutions Canadian farmers and ranchers need.Olds College needs first to stop listening to synthesized American influencers. Second, it needs to start listening to real Canadian farmers. And third, Olds College needs to stop listening and reacting and start leading.