I’m Caleb Harper, principal investigator and director of the Open Agriculture initiative at the MIT Media Lab. Kent Larson courtesy of MIT Media Lab

In his book Tomatoland: How Modern Industrial Agriculture Destroyed Our Most Alluring Fruit, Barry Estabrook details how grocery store tomatoes are both less nutritious and delicious than those grown decades ago. Industrial farming now grows crops for yield, sacrificing taste and vitamins for an easy-to-harvest, shippable product. It’s why apples at your local supermarket are probably about a year old. Caleb Harper, a principal research scientist at MIT and director of the OpenAg Initiative, wants to use technology to grow food that’s healthier, tastier, and more sustainable.

“Growing for nutrition and growing for flavor, it’s not really something anyone does,” he told Digital Trends at the recent ReThink Food conference in Napa, California.

With his background in architecture, Harper may not seem like the most obvious choice for the role of food system shaker-upper. Starting in 2014 with a couple of Dixie cups for his plants, he tried to learn the basic mechanics of what makes them grow better. Four years later, Harper has created the Personal Food Computer to help control all aspects of climate in a cubic foot of space. The boxes are outfitted with sensors, cameras, and circuit boards. Harper and his fellow nerd farmers, as he calls them, experiment with different ways to grow basil or a head of lettuce and then share the results. Everything is open source.


There are commercial smart gardens on the market, such as the Click and Grow or the GroBox. However, these devices are meant to grow a small amount of food with virtually no input from the owner. The manufacturers have dialed in the parameters for growing strawberries or basil. The Personal Food Computer, on the contrary, is meant to be tinkered with.

“Because it has 10,000 years of evolutionary history, the plant doesn’t care if it tastes good,” Harper said. “The plant doesn’t care if it gives nutrition to a human.“

Nerd farmers, though, can run experiments to see if a certain temperature ranges lead to crispier apples or which soils promote more flavorful broccoli. To get to the most appealing taste, Harper imagines a rating system, like Yelp or Amazon. Perhaps regional preferences will emerge, with certain areas of the world opting for a more citrusy basil and others gravitating to a sweeter variety.

“Could we see the corn belt as 3,000 individual microclimates?”

Companies such as Ferrero and Target sponsor Harper’s research. For Ferrero, maker of Nutella, he created a tree computer. In the cube, he’ll recreate 30 different climates, seeing how the trees grow and produce the highest-quality hazelnuts that are not only tasty but use the least amount of water, fertilizer, and pesticide.

“Could we look at the corn belt and say, instead of all corn, which may or may not be good for nutrition and for other things, could we see it as 3,000 individual microclimates?” Harper said.


By controlling climate variables such as carbon dioxide, air temperature, humidity, and many others, OpenAG’s Personal Food Computer allows you to fine-tune and manipulate the conditions to find out what makes the best-tasting, most nutritious plant. Better yet, you can build your own. OpenAG

The area of the world with a matching climate might be a place that’s never grown hazelnut trees before, he said. It’s an idea he calls climate prospecting. Instead of trying to make snow peas thrive in a cabbage climate, farmers would know which crops would do best in their neck of the woods with least amount of resources.


In the 1760s, Carl Linnaeus took the exact opposite approach, trying to grow tea plants in his native Sweden, so the country wouldn’t have to import it from China any longer. The famed botanist wasn’t concerned about invasive species overtaking native ones, but it is something Harper is aware of.

“We’re literally trying to create things like diabetes basil or heart disease basil or degenerative brain disease basil.”

“I think a lot about if I was to plant something or encourage the planting of something that destroyed an ecosystem, we certainly don’t want to do that,” he said. “But to not do that we have to get so much more information about how to go and where to grow things and why we’re doing it there in the first place.”

Cacao. Bananas. Beer. Every few months, there are stories of a different crop in trouble, either from disease, climate change, or both. Yet plants are incredibly resilient. They must be.

“Predators come to them. Bad climate comes to them. Diseases and bacteria come to them,” Harper said. “They have all these — it’s a Swiss army knife of survival.”

Whereas bacteria can easily develop resistance to single-molecule drugs, plants have multiple lines of defense against disease. With that in mind, Harper hopes to grow fruits and vegetables that will help combat chronic conditions before they start.

“We’re literally trying to create things like diabetes basil or heart disease basal or degenerative brain disease basil,” he said. But as anyone who’s tried to figure out if coffee is good or bad for you knows, this is incredibly complicated. What proteins would make one basil better for heart disease or diabetes? It’s one reason Harper is working with Northeastern University’s Barabási Lab on the Foodome project. If you think what makes up a hot dog is one of food’s greatest mysteries, just take a look at the makeup of garlic. According to the USDA, raw garlic is made up of dozens of chemical compounds. The Foodome project puts that number at over 2,000.

“There is a huge difference between the Canadian carrot and a carrot in California.”

Those missing amino acids and nutrients are what Peter Ruppert, a consultant at the Barabási Lab, calls the dark matter of nutrition. They haven’t been quantified, so it’s unclear how the human body metabolizes them and what effects they have on health. What the lab needs is a giant database of food, broken down to the elements. But even that’s not straightforward.

“There is a huge difference between the Canadian carrot and a carrot in California and a carrot in South Africa, because there are different chemicals in the soil, different pesticides,” Ruppert said.

Boil one Canadian carrot and bake another, and those end up different. The lab applied for a grant to get funding to travel to the 50 states and perform mass spectrometry on a plant, perhaps garlic or basil, from local farmers in each one. For now, researchers are gathering data from institutions such as the Alberta Food Composition Database, the source for garlic’s 2,078 compounds.


The other piece is data mining literature around the health effects of chemical compounds found in these foods. A study from the Francis Crick Institute, for example, found that the compound indole-3-carbinol may protect against certain kinds of cancer. That’s a compound found in broccoli, cabbage, and kale. By cross-referencing the two data sets — the compounds in food and the ones that have health benefits — the Barabási Lab can make an educated guess about what proteins Harper’s team should focus on for its diabetes basil.

Part of the challenge will be figuring out how something like indole-3-carbinol interacts with other compounds and how they affect its ability to control inflammation. If it all sounds complicated, think about how difficult it is to get someone to permanently change their diet. “Instead of just pushing people to eat healthier, I can create options which taste the same, give you the same feeling, and on the other hand can be better for your health,” Ruppert said.


Open Agriculture initiative

Harper envisions a future where grocery stores stock regular tomatoes next to more expensive ones that are specifically high in lycopene that also taste like they were grown in your neighbor’s garden. As DNA sequencing becomes more inexpensive, it’s not a far leap to think about the day when it’s not just diabetes basil, but the Smith family’s diabetes basil.

It’s not a far leap to think about the day when it’s not just diabetes basil, but the Smith family’s diabetes basil.

“My grocery store is not going to stock vegetables just for me,” Harper admits, “and so I’m going to grow these few things. And by the way, I’m going to be prescribed certain bacteria that go into the system to make that system colonize a very special product for myself.”

You’ll send your profile to the local restaurant, and they’ll be able to design a steak for you. “Especially if it’s a lab-grown steak,” Ruppert said.

For those of us who think lycopene sounds like a chemical compound that protects against werewolfism, a more transparent food system might be a bit intimidating. It would also require profound changes in agriculture, the supply chain, and healthcare. Doctors may one day prescribe specific bacteria or proportions of chemical compounds, but we’ll still have to rely on the food industry in many ways. That requires trust. “I hate to say Internet of food, because it’s way overplayed in my scene,” said Harper, but he sees a parallel in a decentralized, democratized Internet. To get there, food companies will have to embrace his open-source ethos.

“I really hope that we force transparency and we force trust as our main issue,” he said.

Via Digitaltrends