Last week in the pampas of South America, I stood with a farmer in a field of soybeans that stretched to each horizon. As he absent-mindedly chewed on dry beans to test their moisture content we talked about the future of agriculture. What role will protein-rich soybeans play in meeting the 70% increase in demand for food over the next 30 years? Will breeders develop new varieties fast enough to tackle pests, pathogens and climate change? Is there a way to use fewer harmful chemicals on the land we use to grow food?
For the last decade I’ve been having similar conversations with growers and breeders from the Dakotas to New Zealand. What’s changed recently is their sense of urgency and an awareness that current tools aren’t equipping farmers to face these challenges.
The X team talking with a strawberry grower in California
Agriculture is an incredibly complex field (pun intended!). Farmers make hundreds of decisions every season about what to plant, when to plant, how much insurance to buy and how much to fertilize. They also have to integrate variables like the weather, soil quality, market conditions and many other unknowns to make high-stakes decisions that can make the difference between profit or loss for a farm.
While many farmers use digital tools to help — sensors, spreadsheets and GPS have replaced pencils, notebooks and steady hands — many tell us that these new streams of data are either overwhelming or don’t measure up to the complexity of agriculture, so they defer back to things like tradition, instinct or habit. For these reasons the industry remains one of the least digitized.
I believe we’re entering a new era that we’ll come to think of as “computational agriculture”, and that some of the challenges farmers face today could be helped with a mix of better data, machine learning, and yet-to-be developed technologies. Computers can crunch vastly more data than humans and are really good at working on complex problems with multiple variables and dependencies. These capabilities have led to breakthroughs like Google reducing its data center energy usage by 40% in a field where human experts believed further efficiency at that scale wasn’t possible. Similarly, new computing techniques could help farmers find opportunities in their operations to reduce their use of harmful chemicals or make better decisions about crop-threatening issues like pests, diseases or drought.
In the coming years we envision new tools that will enable farmers to start asking entirely new kinds of questions: What if farmers could manage the plant instead of the plot? What if crops could be bred 10X faster, 10X cheaper? What if any farmer could have access to the best advice anywhere, rather than being limited to personal or local know-how?
Early X-plorations in the field
X’s secret to taking moonshots has always been to look for radical new approaches to problems, whether it’s self-driving cars or balloon-powered Internet — and as Astro Teller said at the World Agri-Tech Innovation Summit in San Francisco today, we suspect at least a few out-there, seemingly impossible ideas are going to be needed to feed the world sustainably and rebuild the planet’s soils. It will also take a new ecosystem of individuals and organizations working together in surprising new ways to make this happen. If you have a piece of this puzzle, and share our excitement about a world where everyone has access to nutritious food, made in a way that’s also good for the planet, please get in touch here.