Green shoots, AI roots: Responding to crises and building sustainable futures with gen AI
Ana Vidal
Contributing Writer
Matt A.V. Chaban
Senior Editor, Transform
Nonprofits are building AI models and agents to tackle global challenges — and uncovering business lessons for organizations everywhere.
There are many things that can play a role in crisis response, whether that’s foil blankets, hardy volunteers, farflung donations, or scientific breakthroughs. Now, generative AI is poised to find a place in the emergency preparedness kit. As the world confronts a long and growing list of challenges — wars, housing shortages, misinformation, climate change, and the increasing natural disasters, migration patterns, and economic hurdles that many of these bring with them — humanity is once again turning to technology to help fix things.
For all the enthusiasm around generative AI to help tackle these crises, it really is striking just how quickly it’s begun to have a real impact not just in business but all corners of life, from the desperate to the sublime. Just consider the recognition this month of not one but two Nobel Prizes for work on AI. As Demis Hassabis, head of Google DeepMind and winner of those awards said, “I’ve dedicated my career to advancing AI because of its unparalleled potential to improve the lives of billions of people.”
Perhaps no issue confronts more people than the climate change sweeping the globe — and it is often cited as one of the areas where AI could hold a great deal of promise in coming up with novel or unexpected solutions to what’s so far seemed a runaway problem. At the same time, AI is showing its potential in addressing local issues, whether that’s disasters and migrations that have grown due to climatic events, or the all too-common dislocations of people for economic, political or social reasons.
These are challenges far from business, and yet they regularly weigh on it. As a constellation of nonprofits and NGOs have begun embracing AI to help tackle global and local issues, their work can provide lessons and insights for ways this emerging technology could have benefits to workplaces and workforces, as well.
This is the second post in our series looking at Google.org’s inaugural Accelerator: Gen AI program and the organizations who are using generative AI and other machine learning to tackle the world’s problems (applications for the next cohort are already open here).
Having explored the potential to address economic opportunity, now we look to crisis response, and the way AI is helping to meet climate, migrant, and global development goals. Their work demonstrates the unexpected ways generative AI is sparking innovative research, up-to-the-minute coordination and communication, and deep analytical insights — AI applications that are relevant to most every organization, not just nonprofits. We get to hear first hand from the leaders of these organizations just how they’re doing it, too.
New frontiers in sustainable materials
To achieve the United Nations Sustainable Development Goals by 2030 and contain damaging temperature increases, global cooperation and innovative solutions are required — but often come at a high economic cost. One area where gen AI could play an important role is providing quick access to crucial information, helping us make the most of our resources, and streamlining processes.
AI is also a key player in the shift to renewables like solar and wind. It helps smoothly integrate these sources into the power grid, balancing supply and demand for a cleaner, more efficient energy system. In short, AI can help the world regulate resources and reduce waste — something the planet desperately needs.
Generative AI is also a driving force behind breakthroughs in critical areas like climate science, medicine, space exploration, materials sciences. One such innovator is Materiom, a UK-based NGO building a database of bio-based materials that can replace plastic and other non-renewable products. The organization estimates that gen AI can accelerate the discovery of sustainable materials nine-fold, allowing entrepreneurs to create eco-friendly products faster and at a lower cost.
As part of its work with the Google.org accelerator, Materiom developed a gen-AI tool enabling anyone, anywhere to create compostable, bio-based materials. This empowers industries like packaging, textiles, and construction to move away from resource-intensive practices, shifting toward more sustainable, eco-friendly solutions. You can explore their work on new materials here.
AI is an amazing set of tools that we can use not only to accelerate these discoveries but also to collaborate and use the knowledge that we already built as a human civilization to its best purpose.
Alysia Garmulewicz, Founder and co-CEO, Materiom
In business: Materiom’s work on identifying new sustainable materials is — like DeepMind’s Nobel-winning work on protein identification and drug discovery — part of a growing list showcasing the dramatic advances AI is spurring in mature fields. Not only will companies be able to directly benefit from using sustainable materials like those being discovered by Materiom and its partners, they will also be able to borrow its research techniques and forge their own breakthroughs in most every field.
Urgent information for urgent situations
As of June this year, the United Nations estimates there were more than 120 million people displaced from their home countries globally, or roughly 1 in every 69 people. Nearly a decade ago, the International Rescue Committee launched Refugee.info to provide timely information to those displaced in Greece due to the war of Syria. In 2018, the IRC and Mercy Corp partnered with Google and others to turn this service into a full-on digital platform called Signpost for supporting millions across the globe. The mission of Signpost is to provide people with accurate, accessible and timely information for people in times of crisis.
The IRC team spent six months this year in the Google.org accelerator developing an AI-powered public infrastructure for humanitarian response in 30 countries. Their solution can create hundreds of AI agents that ingest more than 50,000 documents in the Signpost database to enable the rapid deployment of information services in a crisis.
Working in text and voice in native languages in a straightforward, conversational style, these chatbots can deliver critical, life-saving information to those looking for services, stability, and rebuilding during disasters, armed conflicts, and other forms of exclusion. The IRC team has built systems to assure the safety and accuracy of the content for its clients.
We learned that creating programs with mixed forms of information — with both mass and personalized approaches — is the best way to create empowerment with people on complex topics. By leveraging generative AI, we are able to deliver our extensive content of approximately 50,000 documents in a personalized manner tailored to individual needs.
Andre Heller Peracher, Director, SignPost Project, IRC
In business: Chatbots have become one of the most popular applications of generative AI in the short time the technology has been widely available. As the IRC has shown, the applications can go well beyond customer service to meet the very dire or immediate needs of those in crisis with Responsible AI approaches. The adaptability of gen AI language models mean that it can even be tuned to respond differently to different users or circumstances — an approach that can be quite valuable when dealing with a customer who may be frustrated or excited by their interactions with a chatbot.
Insights in seconds, not weeks
In the fast-paced world of scientific and academic research, time and resources can make all the difference. That's why the cost of scientific research — including skilled personnel, infrastructure, materials, and software — can significantly hinder progress. In 2021 alone, research and development costs surged by $72 billion. The cost reduction is something that the arrival of generative AI could radically change for the better, addressing a persistent challenge that even the most accomplished scientists and researchers grapple with.
Far from replacing researchers and data scientists, gen AI has the potential to empower them and amplify their unique expertise, whether that’s through automating the analysis of vast and complex datasets or uncovering patterns and correlations at remarkable speed. With the development sector facing limited funds and massive global needs, The World Bank is always looking for ways to maximize the impact of its programming as well as the investments of other organizations. In collaboration with 189 countries and various partners, the World Bank formed its Development Impact group, known as DIME, to pool research on best practices in development aimed at such goals as eliminating poverty and world hunger.
DIME, as part of the Google.org AI accelerator cohort, has created a generative AI tool that extracts crucial insights from new and historic research literature. The goal is to efficiently provide guidance and improve decision-making for policymakers and development researchers working to improve outcomes from aid and loans. ImpactAI, DIME’s AI assistant, is aimed at making connections in the literature for areas like poverty reduction, public health, food security, and energy efficiency.
With the ability to provide answers 99% faster through its web interface, this innovative tool significantly improves the efficiency of research and analysis, helping to accelerate critical decision-making processes.
Historically, doing manual curation of evidence would take weeks if not more — it was really time consuming. By automating the process with gen AI, it takes just a couple of seconds to scan through the literature to get the insights needed to make the right decisions.
Samuel Fraiberger, Data Scientist and DIME AI Program Lead, The World Bank
In business: DIME’s work exemplifies how gen AI can optimize resource allocation, enhance decision-making, and drive meaningful change across sectors. We are awash in data these days, so it can be challenging to filter out the noise. The World Bank shows how leveraging gen AI to drastically reduce the time needed to extract relevant insights from research literature — or company data, or any other large pool of information — can lead to both faster and better decision-making for leaders of all organizations, doing so in a matter of seconds rather than weeks.
