Exclusive: Alphabet X is exploring new ways to use AI in food production

Alphabet’s X, the secretive lab charged with finding radical “moonshot” solutions to some of the world’s biggest problems, is exploring ways in which AI could dramatically improve food production. Astro Teller, the head of X, revealed the plan at MIT Technology Review’s annual EmTech Digital event in San Francisco.  

Teller declined to give specific examples, saying X’s team hadn’t yet zeroed in on particular approaches, but he hinted it was looking at how machine learning could be combined with advances in areas like drones and robotics to advance farming practices. 

To be worthy of X’s attention, a project must fulfill three criteria: it has to potentially solve a problem that affects millions or billions of people; it has to involve an audacious, sci-fi sounding technology; and there has to be at least a glimmer of hope it’s achievable within a five-to-ten-year timeframe.

AI-driven agriculture seems a perfect fit. The Food and Agriculture Organization of the United Nations estimates that 20-40% of global crop yields are lost each year to pests and diseases, despite the application of millions of tons of pesticides, so finding better ways to create more productive and sustainable farming will benefit billions of people. As our coverage from this week’s conference shows, AI is certainly an audacious technology, and the speed at which it’s advancing means X’s timeframe feels achievable.

Discussing the challenge ahead of his presentation, Teller said that farming is a hugely complex undertaking and so X isn’t taking it lightly. He cited several examples of where machine learning could potentially help farmers. One would be using the tech to amplify their expertise in crucial areas like deciding when to harvest crops or to apply irrigation.

Another would be to plan their farming in an environment where climate change is making it ever harder to predict weather patterns and how things pests and other factors can affect production.

Asked if this will involve drones, ground-based robots or a combination of the two, Teller said it’s really early, though he noted it helps to have machines very close to what they’re looking at.

X has been looking into agriculture for quite some time. A few years ago, it launched a project on vertical farming, which involves growing crops indoors on racks that stack on top of one another. It made some progress in automated harvesting and other areas, but eventually shut down the initiative because it could not grow staple crops.

Also, there are plenty of startups and bigger companies experimenting with things like using robots to harvest things like strawberries and to gather data about crops via drones. Still, with the brainpower—and AI power—available at Alphabet and X itself, it should be able to harvest some innovate new ideas to transform farming.