Creating an Online Toolbox for Understanding and Communicating Artificial Intelligence within Sustainable Agriculture

Two videos were created to cover introducing and defining AI. In the introductory video, Dr. Alex Kirkpatrick introduces the 10-part video series designed for agricultural professionals. They explain that the course serves as a toolbox for understanding and communicating about agricultural AI. The video outlines the course's dual focus on technology and the human element of AI. 

In the second video of the series, “Defining Artificial Intelligence,” Kirkpatrick provides a foundational framework for understanding what constitutes artificial intelligence. They define AI using its U.S. State Department definition and shorthand definition that serves to clarify the concept. To illustrate these definitions, the video compares various technologies that people encounter in both personal and professional life. Kirkpatrick continues to provide concrete examples of what AI is and is not. The video concludes by encouraging agricultural educators to ground their outreach in common examples of AI to help their audiences realize that they are likely already AI adopters. 

In the third and fourth videos of the series, Kirkpatrick distinguishes between the theoretical possibilities of AI and the practical “Weak AI” currently used in agriculture. They explain that while concepts like Artificial General Intelligence and Superintelligence dominate the public imagination, but the industry actually relies on Artificial Narrow Intelligence, which is designed for specific tasks like autonomous navigation or disease detection. Kirkpatrick encourages educators to focus on these tangible, narrow-function tools to foster more productive and less speculative dialogues with agricultural audiences. 

In the fourth video, “Machine Learning and Other Concepts,” Kirkpatrick decodes essential AI terminology, focusing on the mechanics of Large Language Models (LLMs), machine learning, and deep learning within an agricultural context. A major highlight is the use of digital twins—virtual models of physical farms that allow land managers to simulate stressors and management decisions without real-world risk. Kirkpatrick also addresses critical challenges of AI in agriculture, including the environmental energy cost of training these models and the mysterious nature of their decision-making processes. 

The fifth video of the series, “The ‘4th Industrial Revolution’,” addresses these objectives. This video frames AI as the centerpiece of the Fourth Industrial Revolution, a global shift blurring the lines between physical, digital, and biological spheres. Kirkpatrick explores the “AI arms race,” contrasting the United States' free-market, innovation-first approach with China's state-driven military and economic ambitions—noting that China produced three times as many AI patents as the US in 2022—while the European Union positions itself as the primary global regulator for ethical oversight.  

In an agricultural context, the video illustrates how these technologies can drive sustainability through precision water and nutrient management, autonomous smart farm operations, and other examples. Ultimately, Kirkpatrick emphasizes that while AI offers immense potential for productivity and waste reduction, it also has a high environmental cost and could impact the workforce. 

The sixth and seventh videos of the series cover potential risks and threats of AI. Kirkpatrick outlines critical technical vulnerabilities, such as model drift and skill fade. The lecture addresses social and ethical challenges, including algorithmic bias in lending and the AI divide. Fostering trust in agricultural AI requires experts to move beyond raw data and validate the perceived risks and privacy concerns of ag workers. 

In the following video, “Public Sensemaking of Risk and Threat,” Kirkpatrick explores empirical strategies for communicating the risks of AI to agricultural audiences, primarily through the lens of the Extended Parallel Process Model. Kirkpatrick also introduces the Risk Information Seeking and Processing model. To build the trust necessary for conversations about AI risk, they advocate for transparency, active listening, and the use of emotive, personable messaging that demystifies AI by highlighting its everyday presence. Finally, the lecture emphasizes that understanding an audience’s specific media diet and lived experience is vital for translating technical risks into actionable knowledge. 

In the eighth video of the series, entitled “Public Understanding of AI,” Dr. Kirkpatrick explores public understanding of AI. They highlight the gap between the ubiquity of artificial intelligence and public awareness, noting that 45% of Americans believe they rarely use AI. Kirkpatrick highlights that more Americans are concerned than excited about AI, particularly people of color, younger workers, and those with lower formal education. Kirkpatrick argues that because agricultural audiences are understudied and diverse, communicators must proactively analyze their media diets and use everyday examples to ground discussions, ensuring that strategic framing is used to bridge the gap between technological reality and public perception. 

In the ninth video of the series, entitled “Technology Adoption,” Kirkpatrick examines technology adoption in agriculture, emphasizing that strategic communication must be tailored to an audience's specific stage on the technology adoption curve. On an individual level, Kirkpatrick argues that adoption is driven by four key perceptions: the expected gain in performance, the anticipated effort required, the influence of social peers, and the existence of facilitating conditions like technical support and infrastructure. By identifying an audience's adopter category and addressing these psychological drivers, Kirkpatrick posits that communicators can translate complex AI innovations into practical agricultural applications. 

The tenth video of the series, “Trustworthy AI and Trustworthy Communication,” explores the role of trust and trustworthiness in science and environmental communication, particularly regarding the adoption of AI in agriculture. They argue that socioeconomic status, political ideology, and trust are powerful drivers of attitude and behavior. Kirkpatrick delineates three dimensions of human trustworthiness (benevolence, openness, and integrity) and six technical considerations for AI trustworthiness (reliability, predictability, accountability, performance, safety, and data security). They advocate for interactive communication strategies. The lecture explains that fostering mutual trust is a more effective and achievable goal than knowledge transfer alone. 

The final video of the series, “AI Ethics,” examines the ethical landscape of AI in agriculture, focusing on the moral responsibility, transparency, and socioeconomic impacts of autonomous systems. Kirkpatrick positions agricultural professionals as ethical stewards who must work with developers and producers, advocating for systems that are safe, reliable, and explainable.