Since the early 1980’s, when personal computers first became both available and relatively affordable, farmers wishing to gather and analyze additional information about their farming operations had to make significant investments to bring such functionality into their operation. As a result, the willingness and ability of farmers to take advantage of such opportunities has typically varied considerably by scale, with early investments more common among large-scale farming operations and only later (if at all) accessible to small- and medium-sized farming operations.
Various studies indicate that farmers adopted this new type of technology relatively slowly at the beginning. My doctoral dissertation addressed this topic, finding that only about 4 percent of Iowa farmers had purchased their own computer based on a 1982-84 survey conducted by the Department of Journalism and Mass Communications at Iowa State University. The rates of adoption among firms in the U.S. agricultural supply chain were initially much greater. A survey of New Mexico agribusinesses in 1987 found that 44 percent of respondents had used a microcomputer in their business. Among farmers, the early uses of their computers included financial recordkeeping and for dairy operations, tracking the daily output of individual cows in their herds. By 2021, it was estimated that between 69 and 73 percent of U.S. farmers had access to a laptop or desktop computer, lagging between 20 and 25 percent behind computer ownership rates among U.S. households in general.
More recently, we have seen a similar divide with respect to farmers’ willingness to invest in equipment that allows them to receive and evaluate field level data about the performance of their crops. During the 1960’s, farmers with large scale operations had the financial wherewithal to commission photography firms to provide aerial photographs of their land, but it was an expensive activity and rarely undertaken more than a few times per year. Farmers did not perceive images taken by satellites to be of value to them until the late-1990’s, when the U.S. government removed restrictions on the resolution quality of the images that could be provided to the public.
Today, row crop farmers can install receivers on their farm equipment which allow them to receive signals from GPS satellites in orbit in real time while cultivating their fields, which provides information about the performance of their crops down to the square centimeter level. Horticultural crop and livestock producers can collect similarly detailed information about their operations from miniscule sensors embedded within individual plants or animals. However, the equipment needed to both receive and act upon the detailed information is expensive, so farmers with large operations are much more likely to purchase such equipment than their small- or medium-sized counterparts. According to data collected regularly through USDA’s Agricultural Risk Management Survey (ARMS), equipment such as auto-steering and yield monitors has been adopted by 70 and 68 percent of large-scale U.S. farmers respectively, but only by 9 and 13 percent of small-scale farmers.
The latest offering in this area is the public availability of artificial intelligence (AI) models for use by both households and businesses. One major difference between this innovation and those previously discussed is that deploying AI to help farmers collate and analyze information about their operations does not necessarily require the same significant level of investment in capital equipment as did the earlier generations of information technology. Thus, in recent years, farmers in both developed and developing countries have been able to utilize AI technology to help in their decision-making processes.
In developed countries like the United States, this process involves analyzing data gathered through precision agriculture tools such as yield monitors and IOT sensors monitoring plant growth and soil moisture that was already being collected on many farms prior to the deployment of AI models in the 2010’s. For example, in the state of California, by far the largest producer of horticultural crops in the United States, farmers growing an estimated 65 percent of those fruit and vegetable crops (as of 2017) rely on irrigation for those crops to thrive. At the University of California at Merced, Professor Wan Du has developed a system using AI modeling and data collected from IOT sensors placed within orchards to optimize delivery of irrigation water to tree crops like almonds and avocados.
Under a grant from USDA’s National Institute of Food and Agriculture (NIFA), scientists and extension staff at the University of California at Davis developed a new tool (called Leaf Monitor) that allows farmers to image the leaves of their crops using a handheld spectrometer. Those images are then uploaded to an AI machine learning system to evaluate the leaf traits and their nutrient levels, using the information to build a predictive model for the crop whose images the farmers submitted. This model identifies nutrient deficiencies and prescribes treatments in terms of appropriate fertilizer application rates.
The non-profit organization Digital Green has developed an AI-based app called Farmer.Chat which it made available to small-holder farmers in Kenya, Nigeria, Ethiopia, India, and Brazil, starting in late 2023. This app is designed to provide localized solutions to problems experienced by crop and poultry farmers in those countries, responding to farmer queries inputted through voice, text, and image-based searches. During the summer of 2025, Digital Green commissioned a survey of Kenyan farmers who had used the app previously to gauge their impressions of the information they gained from their past queries from an organization called 60 Decibels. The survey had 450 usable responses, from a population that was 72 percent male and 28 percent female.
In general, the app was very well received by the respondents. The survey found that a majority of respondents (68 percent) believed that the information they received ‘very much’ improved their way of farming, and more than 80 percent had never had access to such information services before they used Farmer.Chat for the first time. Using this app required no significant new investments by the farmers, as long as they already had a mobile phone.
