But the Bezos Earth Fund is looking at how “modern” AI—more efficient AI tools and technology—can be leveraged to protect climate and nature.
On Thursday, the Earth Fund announced $30 million in grants will be awarded to 15 teams from universities and environmental groups across the world.
Created with a $10 billion commitment by Amazon.com founder Jeff Bezos, the fund has tapped the expertise of Nvidia, Alphabet’s Google, Microsoft subsidiary Microsoft Research, nonprofit AI research institute Ai2, geographic information system (GIS) software developer Esri, and Amazon Web Services, to support groups more adept at understanding the nuances of nature than technology.
“The more that we support AI to be a force for good, that will ultimately be the mechanism by which we save the planet,” Amen Ra Mashariki, director of AI at the fund, said in an interview.
The grant winners include a team at the New York Botanical Garden that is “automating plant species identification with computer vision and AI,” a group from the University of Leeds in England that is creating an AI platform that will convert food waste into microbial protein, and a team at the University of Witwatersrand in Johannesburg that is developing a weather forecasting tool kit for Africa using AI, according to the Earth Fund.
The awards mark the second phase of a multiyear $100 million “AI Grand Challenge” that began in April 2024 with $1.2 million awarded in grants of $50,000 each to 24 groups to develop “AI-ready” proposals, says Mashariki, who was a senior principal scientist at Nvidia before joining the Earth Fund.
The winners announced on Thursday were selected from that initial group, and will each receive $2 million to implement projects addressing biodiversity loss, food insecurity, and climate change over the next two years.
In an interview, Lauren Bennett, program manager of spatial analysis and data science at Esri, says she was drawn to the challenge because it wasn’t about groups using “AI for the sake of it.” Instead it was “trying to help conservation organizations do their jobs.”
Many of those vying for grants already use Esri’s software. “The last thing they are going to do is something irresponsible just to put AI next to it,” she says. The challenge was “inherently set up to avoid all of that because of who they are.”
The Earth Fund’s strategy was to find organizations with projects already under way that could be elevated through the use of AI, rather than creating problems for AI to solve, Mashariki says. That resonated with Esri’s Bennett. “Technology only matters if it’s used to do something useful or important,” she says.
For instance, the Nature Conservancy has been working for some time to curb illegal and improper ocean fishing practices—such as catching endangered or threatened species—by employing electronic monitoring technology.
With its Earth Fund grant, the organization plans to employ edge AI, that is, AI technology that captures data near where it was created on local devices rather than in the cloud. Using Nvidia Jetson sensors on fishing boats, catches can be tracked as they are brought onto a boat, boosting the scale and efficiency of the nonprofit’s work.
“Those are the types of problems that we’re excited to see because there is a clear baseline and a clear expectation of the value proposition of using AI,” Mashariki says. “Speed, scale, accuracy, precision, and efficiency is what we’re looking for.”
Tech companies have been involved with the Earth Fund challenge since phase I, sharing knowledge of how AI can be used to solve problems and directly mentoring teams as they developed their proposals. Esri, for instance, held a “technology transfer” session exploring the full capability of its software, Kevin Butler, a product engineer on Esri’s analysis and geoprocessing team, said in an interview.
“In that session and subsequent individual meetings, we learned about [the grant seekers’] approaches and were able to ask individual questions and explore how a spatial approach might enhance the story that they could tell or the type of analytics that they might do,” Butler says.
Support from the participating tech companies meant providing free technology for some and/or access to their professional services.
Mashariki’s AI team at the fund will also be available to support the grant winners, as will program directors in topics such as biodiversity and climate change. “I [have] deliberately used words like, ‘we’re excited to support you, partner with you,’ because we’re not just funding them,” he says.
