Agriculture’s importance means it should get the best tech
- Elliott Grant, Ph.D., is the chief executive officer at Mineral, an Alphabet company.
In November at COP27, world leaders gathered at Sharm El Sheikh and bemoaned the impact agriculture is having on the world’s climate. Collectively, food production contributes about a quarter of the world’s greenhouse gas emissions. Bold initiatives to eat less meat, protect biodiversity, and sequester carbon in farmland were announced.
But however worthwhile and earnest these efforts are, they overlook a hard reality. A changing climate, supply interruptions and inflation, pests and disease resistance, regulatory requirements, out-of-date incentives, new practices such as cover cropping — all are increasing complexity and risk for farmers. The problem is not that farmers don’t care about the environment. In fact, I know they care a great deal as it supports their livelihood, and farmland is often their greatest asset. The problem is that farmers don’t have adequate tools to produce more, while using less, under increasingly challenging conditions.
If agriculture is so important, why isn’t it getting the best tech?
Despite the hype, private and public investment in agriculture technology is far, far smaller than investments in other climate solutions, such as clean energy (for example, venture capital and private equity invested almost 70 times more in cleantech than agtech in 2021). If we are serious about helping agribusinesses transform and adapt, then they need access to the latest technologies and best science, adapted to the immense challenges of the global food system. Take data, for example. As an industry, farming remains one of the least digitized, according to McKinsey & Co. This poses a profound problem in our efforts to address the large-scale challenges facing agriculture today.
In the past couple of years, software engineers have seen exciting results from decades of research into artificial intelligence, or AI. Recent examples are so-called “large language models” such as GPT or LaMDA that can hold a cogent conversation with a human, generate software code, or write essays in response to a prompt. Image recognition and computer reasoning are good enough to enable cars to drive themselves on city streets. Algorithms can predict the complex shapes of proteins. Some of these programs can generate patterns from data that lead to new knowledge and understanding. The issue is that these computer models needed considerable amounts of high-quality data to become useful. If agriculture isn’t sufficiently digitized, it will miss the AI train.
Yet I am optimistic. Smart equipment that senses every plant, drones that can fly autonomously and spot diseases, and cheap sensors that can monitor soil are all coming. Within the decade, advances like these will increase the data available to farmers by orders of magnitude. The industry is working on standards to improve farm data security and manage data use. Farmers have existing networks — such as cooperatives, extension services, and 501(c)(5) nonprofits like the Iowa Soybean Association — that can act as agents of change, educate, and ensure the interests of individual farmers are protected.
That trend is why Alphabet incubated Mineral at X, its moonshot factory, over the past 5 years. During that time we sought to validate that the technology had matured to be market-ready, and the opportunity to create a standalone company in agriculture was real. As of Jan. 1, 2023, as a new Alphabet company, Mineral is applying the latest breakthroughs in artificial intelligence and computer science to the existential challenge of helping farmers around the world increase land productivity and sustainability.
We have spent years testing and experimenting on farms around the world — and are eager to now put these tools in the hands of more farmers, to help them achieve the goal of producing 70% more over the next 30 years using less land and fewer inputs.
Ask the plants of the earth, and they will teach you. — Job 12:8
It isn’t about creating a new gizmo or a clever app. Mineral is a foundational effort to bring farmers, crop breeders, advisors, and researchers a radically new way of measuring the plant world and — together with their human intuition — gaining a better, deeper understanding.
Elliott Grant, Ph.D., is the chief executive officer at Mineral, an Alphabet company applying the latest breakthroughs in machine learning, artificial intelligence, and perception technology to make agriculture more sustainable.