Of all the things I did when I visited Cornell University recently, I probably had the most fun brushing up on how plants have sex.

Cornell is one of the world’s top universities for research on improving crops. Their work involves a lot of plant breeding. During one meeting, I got to try my hand at cross-pollinating wheat, which is a surprisingly delicate procedure. It gave me even more respect for the people who do it every day.

Cornell’s work on crop improvement also involves a lot of cutting-edge genetics. You might see the words “crop improvement” and “genetics” in the same sentence and think I’m talking about GMOs. Although Melinda and I do support research in that area—we don’t think poor farmers should be denied the choice to use any tools that might benefit them—the work I saw at Cornell is different. It’s focused on how the science of genetics can improve agriculture in other ways. And the advances are really exciting.

I got interested in crop breeding through my work with the Gates Foundation. Because most of the world’s poor people are farmers, helping farmers grow more food is one of the most powerful levers we have for fighting poverty. The faster we can improve crops—making them more nutritious or drought-tolerant, for instance—the faster we can help farmers become more productive.

My main guide was Dr. Ed Buckler, a scientist in his mid 40s who works at Cornell for the U.S. Department of Agriculture (USDA). Over the four hours we spent together I asked Ed dozens of questions (I’ve learned a lot about agriculture, but I’m still a city boy at heart), and he was always quick with an answer. Yet Ed and his colleagues aren’t just experts in their field—they’re also deeply passionate about their work. I can see why: The advances they’re working on will change people’s lives by dramatically accelerating a process that is now slow and laborious.

Here’s how it works today. Suppose you want a variety of corn with a natural resistance to a certain pest. You start by planting as much corn as you can. You wait 8 to 12 weeks for it to grow, and then you take pollen from some of the plants that aren’t infested and use it to pollinate others. If the offspring of those plants is pest-resistant, you’re in luck—your plant won the genetic lottery. If not, you have to start over. Because you’re limited by the growing season, the process can take seven to ten years.

Genetics research will cut that time in half.

Getting there takes three steps. One is to understand the crop’s genetic makeup. Ed took me on a short tour of a lab where machines called sequencers were analyzing DNA from thousands of plants. They were mapping the genes that give each plant its physical traits: its height, color, etc.

The second step is to go into the field and record those physical traits for each individual plant whose genes you’re studying. Cornell researchers are growing hundreds of acres of corn and other crops not far from campus, and they make regular treks out there to collect data. Unfortunately, I didn’t have time for a field trip on this visit.

Finally, you build a computer model that puts the two together—the genetic maps of individual plants, along with the data about their physical traits. Once you have that model, you no longer need to cross two plants and just hope for the best. You can ask the computer, “Out of all the plants I have in my field, which two should I breed in order to produce one that is pest-resistant?” Think of it as a highly sophisticated Match.com for plants.

Cornell and the USDA have already built such a model for some traits in corn; because people in rich countries eat corn, there’s a big market for better varieties. Meanwhile, crops that are eaten mostly by the poor have largely been ignored by scientists. But that’s starting to change.

With support from the British government, our foundation, and others, researchers at Cornell and the USDA are now working on a model for cassava, a root vegetable that’s a staple crop in many tropical regions. Partners in Uganda and Nigeria are growing lots of plants, recording their traits, and sending genetic samples to Cornell for sequencing. When the cassava model is finished, it will help breeders develop new varieties faster than ever. (Incidentally, I’m fascinated by cassava—and you may not know that it is responsible for the fun factor in bubble tea.)

During my visit, I learned about one trait that I had never thought about before: poundability. Over lunch with several graduate students, a Ugandan researcher named Paula Iragaba told me that women in her country do most of the work to turn cassava into flour, and they wish it were easier to process. “Women’s preferences have to be taken into account,” she said. I couldn’t agree more.

Keep in mind, none of the genetics research I saw changes the basics of plant sex. Breeders in the field still have to move pollen from one plant to another, as they have for ages. We’re just getting a lot smarter about helping them pick the best partners. And the result will be phenomenal—a much faster path to more-productive crops so that millions of people can eat better food, earn more money, and improve their lives.