Picture this: Colossal, gas-powered autonomous robots bulldoze across acres of homogeneous farmland under a blackened sky that reeks of pollution. The trees have all been chopped down and there are no animals in sight. Pesticides are sprayed in excess because humans no longer tend to the fields. The machines do their jobs—producing massive amounts of food to feed our growing population—but it's not without ecological cost.
Or, envision another future: Smaller robots cultivate mosaic plots of many different crops, working around the trees, streams, and wildlife of the natural landscape. They’re powered by renewable energy sources, like the sun, wind, or maybe water. Agrochemicals are a thing of the past, because the robots help the ecosystem remain in harmony, so pests and superweeds are kept at bay. It’s a futuristic Garden of Eden, complete with blue skies, green pastures, and clean air.
Which world would you want your food to come from?
These are the two futures imagined by Thomas Daum, an agricultural economist at the University of Hohenheim, who works on food security and sustainable farming in places like Uganda and Bangladesh. In July, he published a thought piece in Trends in Ecology & Evolution that laid out twin visions of an ecological utopia or dystopia in an effort to discuss how the technological revolution in farming—also known as Agriculture 4.0—could shape our future.
“Today’s farming has to change,” says Daum, who worries that the disruptive effects that agricultural technology is having on the environment aren’t getting enough attention. The climate change mitigation strategies outlined in the Paris Agreement cannot be met without transforming how we grow food. “Even if you change all the other sectors,” he says, “if you don’t change agriculture, we will still miss those targets.”
Even in a world without massive farm robots, large-scale farming practices are already changing the environment. “Agriculture inherently is an intentional shaping of the ecology of a particular place,” says Emily Reisman, a human-environment geographer at the University of Buffalo. We remove wildlife, degrade the soil, and clear the land to better grow food, as well as spray chemicals to ward off pests and disease.
When we add existing farm technologies to that mix, well, it gets worse. Machines like tractors, harvesters, and crop-monitoring drones generally require controlled environments to function efficiently, so unpredictable factors must be eliminated as much as possible in industrialized farming. This can mean year after year of monocropping on perfectly level fields with little variation in growth, everything ripening at the same time, and the frequent application of herbicides, pesticides, and fungicides to ensure uniformity. The standardization is a result of our need to mechanize agriculture, says University of Rhode Island sustainable food systems scientist Patrick Baur. “That is farming and the agro-ecosystem and the entire cultivation process being shaped to meet the needs of the machine,” he says.
The environmental consistency needed for industrialized agriculture has substantially contributed to a loss of biodiversity, the variety of plant and animal life necessary to keep ecosystems in balance. Biodiversity protects water quality, moderates global temperatures by trapping carbon in the soil (instead of in the air), and ensures that there are insects to pollinate the crops and natural predators to decrease the presence of pests. “Machines dramatically reduce the diversity of insect life, microbial life, and flora and fauna,” Baur says, because so much of it needs to be cleared away for them to run optimally.
But why do we need machines to produce food? It’s an issue of economics. To keep up with the ever-rising demands of a growing population, agriculture requires more and more labor. Food is also much cheaper than it was in the past, pressuring farmers to produce higher yields at lower profit. As a result, if field laborers make less money and leave the industry for better-paying options, farmers may increasingly turn to mechanization to fill the gap.
Sustaining both the environment and the growing demand for food is a tough balance to strike. Yet with the rise of artificial intelligence and autonomous devices, Daum sees unique potential for agricultural innovations to help us adjust to the changing climate, and to restore biodiversity rather than threaten it. “Could we not reverse some of these trends that have negatively affected the farming landscape?” he asks. Rather than adapt the environment to meet the needs of technology, we could program technology to meet the needs of the environment, Daum argues in his paper. “Smart” robots governed by machine learning might be able to better function in natural, biodiverse systems, because they can mimic the flexible thinking that humans use to assess their surroundings and make decisions.
Robots are already starting to take off in the food sector; they are picking strawberries, harvesting lettuce, pollinating flowers, and even milking cows. Because they work more efficiently, robots can perform eco-friendly tasks that would be uneconomical if they had to be done by hand, like manual weeding, which can reduce the need for spraying chemicals. Daum thinks they could also help with the labor-intensive maintenance of hedgerows, which are natural fences of wild shrubs or trees planted along the perimeter of farmlands; they promote biodiversity by serving as alternative habitats for native plants, insects, and wildlife. And smart robots may even be better suited for intercropping, or growing multiple crops in the same field, a sustainable farm practice that encourages soil health and decreases pests, but is costly and inefficient to do with current technologies.
But rolling out robots that can do these jobs comes with its own challenges. They’d need to be even smarter than the current generation of lettuce-plucking and strawberry-picking robots—which means it’s going to cost more money to engineer them. And Reisman points out that building and maintaining these machines requires resources that drain the environment, too: It would demand the mining of minerals, like copper and lithium, to build them, energy for smart systems that rely on cloud storage, and a way to dispose of the robots, since they will likely have limited life spans.
In fact, she says, Daum’s vision of an ecological utopia may actually turn out to be a dystopia for human workers. “It’s harder for me to see the utopia when it comes to a labor perspective, and how people can have meaningful, gratifying, well-compensated labor in a world of agricultural robots,” she says. At best, robotic innovation may simply not be necessary, because agroecologists have shown that small-scale farms with abundant human labor can be both highly productive and biodiverse. At worst, it could exacerbate socioeconomic problems that are already present in farming and tech-related industries: exploitative practices in mineral mining, a global underclass of workers programming AI algorithms in unfair working conditions, and the displacement of laborers who are unable to find work.
Daum’s paper was somewhat unusual for an academic journal: It used science fiction storytelling as a literary device to explore the bounds of what is possible. And some of his colleagues appreciated this approach as a way to engage the public in a conversation about what the future can look like. “There’s potential here to get to this eco-utopian future,” Baur says. “I liked the paper—it’s a really pithy, easy to digest, very straightforward distillation of what’s at stake.”
Ira Bennett, a chemist turned social scientist at Arizona State University, uses a similar technique to explore emerging technologies outside of the lab: He develops public assessment workshops that present attendees with futuristic scenarios and asks them to make complicated decisions about governance, policy, and the allocation of resources. “Science fiction is a really powerful tool because there’s very little about our future that has not been explored,” he says, and adds that it’s an accessible way to include different people into these conversations, including environmentalists, farmers and farmworkers, robotics manufacturers, tech developers, and really, anyone who eats food. For Bennett, the ultimate goal of this outreach is to determine the type of future people want to see.
That said, while he thinks the paper is a good start, Bennett says it lacks a clear plan for what should happen next. And because it was published in an academic journal, it will likely only be read by scholars, who may have a different idea of utopia than people who actually work in the industry. “If you’re a farmer trying to run 10,000 acres of wheat, the utopia looks like the machine that does all the farming for you, and does it at a cost per bushel that allows you to make more money,” he says. In countries dominated by large-scale farming, like the US and Argentina, Bennett argues, the idyllic vision of smaller robots symbiotically tending to their surroundings just isn’t realistic.
But Daum stresses that some of the elements of his utopian vision are possible, even for industrialized farms that would require massive robots to sustain high yields. What’s most important, he says, is that we pay attention to farming technologies and policies now, or else we may passively slip into his hypothetical dystopian world. “We are at a crossroads. There is a big opportunity to change the way farming is done for the better,” Daum says. “Societies should have a say into which type of future we want to have.”
Update 8-16-2021 3:27 pm: This story was updated to correct Patrick Baur's job title.