Josh Henretig of Microsoft: ‘We Are in an Information Drought When It Comes to the Natural World’
But AI has ‘enormous potential’ to fill the gap, adds the company’s senior director for corporate sustainability
We’re awash in data — except when it comes to the natural world, says Josh Henretig, a senior director for Microsoft who leads the company’s AI for Earth and corporate sustainability programs. And that dearth of data presents a huge obstacle to making evidence-based decisions about land use and other human activities that bear on planetary sustainability.
The answer? Among other things, says Henretig, we need to leverage artificial intelligence in ways that fill data and data analysis gaps related to environmental challenges. Future H2O’s John Sabo interviewed Henretig to find out more about how Microsoft and its partners are using AI to advance sustainability globally.
John Sabo: Data for good, data for change, data for sustainability — all these imply that data can be persuasive and catalyze shifts in behavior and perspective. How does Microsoft make that dynamic happen? What are the elements?
Josh Henretig: Today, we know more than ever about human activity. More than one-quarter of the 7.6+ billion people on Earth post detailed information about their lives on Facebook at least once a month. Nearly one-fifth do so daily. Those data are fed to increasingly powerful algorithms that link people to others, products or information.
Yet we are flying blind when it comes to understanding the natural world. Scientists still struggle to predict the effects of climate change at the resolution of cities or regions, or over timeframes of months or weeks — largely because they don’t have the kinds of data needed to make such predictions, or because they lack the algorithms to convert data into useful information.
“When it comes to our understanding of the natural world, we are in the middle of an information drought, and are practically flying blind in our efforts to steer human society onto a more environmentally sustainable path.”
Josh Henretig
For example, in the United States, the best available data sets on land cover, at a resolution of 30 meters, were last updated nearly 7 years ago. Globally, the picture is much less complete. Yet without accurate information, housing developers, foresters or other land planners can’t make evidence-based decisions about which parcels of land to use for which purposes, and how much to leave untouched.
By combining breakthroughs in data, infrastructure, algorithms and applications, we are already capable of producing computational services such as speech and image recognition at levels beyond human capabilities — solutions that are already changing the way we work, play, and communicate with each other, and our machines.
Now we’re applying Microsoft’s deep investments in artificial intelligence (AI) research and technology to the unprecedented challenges of mitigating and adapting to changing climates, ensuring resilient water supplies, sustainably feeding a population rapidly growing to 10 billion people — all while stemming an ongoing and catastrophic loss of biodiversity.
And this is why in December 2017, our president, Brad Smith, announced AI for Earth, a $50 million, 5-year commitment to deploy Microsoft’s deep investments in AI research and technology in the four key areas of climate change, agriculture, water, and biodiversity conservation.
The program is designed to empower external organizations working in these areas by: providing access to Microsoft’s AI infrastructure; building a dynamic community of individuals and organizations taking an AI approach to sustainability; producing educational programs on how to use Azure’s AI technology for environmental solutions; and making deeper investments in robust, scalable data and model APIs and applications hosted on Azure.
John Sabo: What are some of Microsoft’s most interesting challenges around data for sustainability? How is it overcoming those challenges?
Josh Henretig: One of the significant challenges we’re trying to overcome is that when it comes to our understanding of the natural world, we are in the middle of an information drought, and are practically flying blind in our efforts to steer human society onto a more environmentally sustainable path.
For example, we know that species are going extinct at a rapid pace — at least 1,000 times more quickly than would be expected in the absence of human activities. These extinctions have devastating impacts on the ecosystem services on which humans rely. Species loss wreaks havoc on pollination services, which puts our global agriculture system at risk. Clean water depends on healthy forests and wetlands, themselves the result of a complex web of species.
“Our lack of insight often boils down to a) a lack of data needed to make valid predictions, and b) a lack of algorithms to convert data into useful information. The potential for AI to help monitor, model and respond to environmental challenges is enormous.”
Josh Henretig
In all of these cases, our lack of insight often boils down to a) a lack of data needed to make valid predictions, and b) a lack of algorithms to convert data into useful information. The potential for AI to help monitor, model and respond to environmental challenges is enormous, and we have some exciting partners that we’re working with to leverage AI in new an interesting ways:
- For example, the Chesapeake Conservancy is using the latest high-resolution datasets combined with powerful AI tools to redefine how landscape conservation is approached. Using advanced flow accumulation algorithms developed by leading scientists, the Conservancy is producing maps depicting concentrated surface flow at the parcel scale. When combined with high-resolution land cover datasets, these drainage maps can assist in identifying areas that have the greatest potential to reduce sediment and nutrient loads into adjacent water bodies. By using technology to identify nature-based solutions at the right place, the right scale, the right size and the right time, the Chesapeake Conservancy can ensure that protected forests and wetlands are contributing to the environmental health and economic well-being of the Chesapeake Bay region.
- Another partner of ours, SilviaTerra, is revolutionizing how we inventory forests by combining satellite imagery with machine learning to create the first high-resolution inventory of all 537 million acres of U.S. forests. With support from Microsoft, SilviaTerra is processing 800 terabytes of data at 10x conventional speeds to create detailed maps of forests at a 15-meter resolution. Conservationists, governments and landowners can use these maps to assess their forests and develop sustainable management plans for a fraction of the time and cost of traditional forest surveys.
John Sabo: What are the trends we should be looking for over the next 2–3 years in this field?
Josh Henretig: In order to adequately address climate change and the serious resource challenges we face, we need better data about the health of our planet, with water being a major focus area, alongside things like biodiversity and agriculture.
One of the big things we’re already seeing is the use of machine learning and other forms of AI to accelerate the efforts of people working on the ground on these issues. Already, it has been instrumental in helping researchers collect accurate and current data, enabling them to monitor and react to how ecosystems are changing. Solomon Hsiang’s prediction of climate-related migration and OceanMind’s fight to preserve ocean health by ferreting out illegal fishing around the globe are two great examples.
As AI continues to develop, we’re excited to continue working with great partners to drive more efficient conservation efforts, in service to a healthier planet for humans and animals alike.
John Sabo: Who else do you look to in the sustainability space as a leader in using data for change? Why?
Josh Henretig: In our work with our AI for Earth grantees, one thing we’ve seen time and time again is that it’s often smaller research organizations or academic institutions that are at the forefront of study and disruption in this space. Microsoft is great at building technology to collect and process information at incredible speeds, and we want to help organizations not be limited by lack of time or resources.
“One thing we’ve seen time and time again is that it’s often smaller research organizations or academic institutions that are at the forefront of study and disruption in this space.”
Josh Henretig
We view our grantees, like the Natural Capital Project, led by Stanford professor Gretchen Daily, as sustainability leaders, as well. They’re the people on the front lines who develop really innovative uses for data. The Natural Capital Project is using remote sensing data with machine learning to pinpoint thousands of forgotten dams, enabling better ecosystem conservation in those areas.
We’ve also been impressed with many of these organizations’ focus on driving open-source solutions and helping build APIs, which enables others to enact change on an even greater scale all around the world.
Josh Henretig is a senior director for Microsoft and is responsible for leading the AI for Earth and corporate sustainability programs for the company. Josh oversees the management and implementation of global initiatives that: 1) reduce the energy, carbon, water and waste related impacts of Microsoft’s operations; 2) put Microsoft’s cloud and AI tools in the hands of people and organizations working to solve global environmental challenges; and 3) enable the company to be a leader, collaborator and innovator on key environmental issues.
Josh has been at Microsoft for 16 years and has played a leading role in the design and execution of the company’s global environmental strategy since the program’s inception. Most recently, he helped establish the AI for Earth program, which was funded with $50 million and a 5-year commitment from Microsoft President Brad Smith in December 2017.
Over the years, Josh has contributed to all aspects of Microsoft’s environmental strategy, from establishing the company’s commitment to carbon neutrality and 100% renewable energy through an internal price on carbon, to working with industry partners like Ecolab to create data-driven tools like the Water Risk Monetizer that help businesses to incorporate water-related risks into decision making, to developing (and launching) the AI for Earth program to assist organizations in leveraging AI to monitor, model and manage Earth’s natural systems.
For more information about ASU Future H2O’s work and research on creating opportunities for global water abundance, visit our website and subscribe to our newsletter.
