Microsoft launched the AI for Good Lab in 2018 to harness the transformative energy of AI to sort out international challenges and enhance lives. The info scientists and researchers within the AI for Good Lab are specialists of their fields from all over the world, and their modern work wouldn’t be doable with no international community of a whole lot of organizations who present important subject material experience for the challenges we are attempting to unravel.
On this version of FYAI, a sequence the place we dive deep on AI tendencies with Microsoft leaders, we hear from Juan M. Lavista Ferres, Company Vice President and Chief Knowledge Scientist for Microsoft’s AI for Good Lab, who’s main Microsoft’s efforts to create a collaborative ecosystem and a workforce of devoted knowledge scientists and researchers, area specialists, and organizations worldwide that drives progress towards addressing a number of the world’s most urgent challenges.
On this Q&A, Juan shares his insights on the rising AI for good motion, how Microsoft companions with native organizations to assist resolve region-specific challenges, and measure and scale AI initiatives to drive impression.
You’re Microsoft’s Chief Knowledge Scientist and Director of the Microsoft AI for Good Lab. What does that really appear to be day-to-day?
My journey at Microsoft began all the way in which again in 2009 once I was working randomized managed experiments as a part of Microsoft EXP. Wanting again, a lot of my work then knowledgeable what ultimately matured into our official AI for Good program. One in every of our first AI initiatives centered on higher understanding the causes of sudden toddler dying syndrome. The analysis we produced with our collaborators helped perceive a number of the causes—and in so doing, bolstered our conviction that AI might be an indispensable instrument in enhancing individuals’s lives.
This additionally illuminated a spot that clearly wanted to be crammed. The majority of AI experience on the planet is concentrated within the tech and monetary sectors, and too many organizations engaged on fixing vital societal issues don’t have the sources or technological experience to use AI to their work. We assist fill this hole by offering AI experience and compute capability to convey cutting-edge expertise to our associate organizations.
I’ve a robust bias for motion, so every day I’m centered on how we will leverage AI to assist essentially the most individuals, as quick as doable. This implies I work with our workforce to conduct utilized analysis, search out and interact various new companions, and information the construct out and deployment of AI options. In some ways, that is my dream job. I get to search out methods to make use of knowledge, apply AI, and enhance the world, and accomplish that with a number of the most gifted researchers and scientists.
Our Lab was one of many first within the AI for Good house, which I’m proud to say has develop into a rising motion. The United Nations, as an example, hosts an annual AI for Good convention which we participated on this July in Geneva. It’s no shock that that is catching on. For lots of the issues we’re engaged on, AI isn’t only a resolution, it’s the one resolution.
Partnership is on the coronary heart of the work we do. We imagine that in an effort to make the most of AI’s potential to profit everybody, we should leverage native, on-the-ground experience to co-create options tailor-made for the communities with whom we’re working. We’ve groups in Microsoft workplaces within the United Arab Emirates, South America, and Kenya as a result of it’s vitally vital for us to work alongside native researchers.
One instance of the significance of partnership in observe is our work to detect historic flooding and enhance catastrophe response globally. Floods are among the many deadliest varieties of utmost climate, affecting hundreds of thousands of individuals and inflicting over USD40 billion in injury every year. To deal with this, we developed a mannequin utilizing 10 years of Artificial Aperture Radar (SAR) knowledge, which might detect floods by clouds, at evening, and in distant areas. That is the primary mannequin of its variety. We labored carefully with companions in Kenya, Ethiopia, Spain, and the United Nations to allow real-time response. In Ethiopia, we used the mannequin to detect practically three occasions extra flood-prone areas than current datasets. It confirmed identified flood zones and revealed new ones, providing a extra full view of flood threat throughout the nation. After devastating floods struck Kenya in Could 2024, we used this instrument to estimate that 75,000 hectares—about 2% of all cropland—have been flooded or in danger.
In lots of instances, the worth proposition we’re bringing to the desk isn’t solely elevated accuracy but additionally velocity. If we will use AI to, for instance, measure deforestation within the Amazon in a matter of days slightly than months, governments within the space can reply a lot sooner.
How do you measure the impression of AI for Good initiatives?
The definition of ‘impression’ is exclusive to every venture, however overarchingly, we assess whether or not and to what extent individuals’s lives have been improved by our work. The important thing to attaining that is being guided by subject material specialists who’ve a strong understanding of the challenges to be solved and the info out there to unravel them. Our aim is all the time to assist our companions and advance their missions, whether or not which means forecasting malnutrition with larger precision or enhancing how we monitor giraffe populations to assist conservation efforts.
We additionally wish to develop the advantages of this expertise to communities that in any other case won’t be capable to make the most of it. For instance, we’re engaged on a instrument to diagnose retinopathy of prematurity in low-resource settings the place pediatric ophthalmologists aren’t out there. Retinopathy of prematurity is an more and more frequent situation, because of enhancements in maternal and toddler healthcare that saves the lives of prematurely born infants across the globe. But when this situation isn’t identified and handled, it will possibly result in lifelong blindness. We’ve developed algorithms that reveal how AI will help detect this situation utilizing movies captured on a smartphone. This has the potential to empower healthcare professionals to intervene early in an toddler’s life and protect their eyesight.
The extra correct and helpful our AI instruments are, the larger their downstream impression will likely be. That is no small feat. AI expertise is evolving and altering quickly, and our researchers work tirelessly to remain on the forefront.
The place do you see the most important want—and alternative—to scale this work?
With a view to actually scale the advantages of AI, we have to make sure that persons are related not simply to AI however to digital expertise usually, which continues to be out of attain for two.6 billion individuals all over the world.
Our workforce has invested closely in creating catastrophe response capabilities utilizing geospatial machine studying and satellite tv for pc imagery. We plan to proceed supporting first responders within the instant aftermath of pure disasters in the USA and throughout the globe. We’ve additionally seen super advantages of utilizing AI for biodiversity conservation, reminiscent of within the Amazon and in Sub-Saharan Africa.
As a non-native English speaker from Uruguay, I take advantage of AI to make sure my grammar in paperwork like analysis papers is pristine and error-free. A overwhelming majority of educational publications and journals are in English. Giant language fashions have made it simpler for myself and numerous others to contribute their analysis, which is a tide that lifts all boats.
What’s a venture you’re most enthusiastic about proper now—and why?
On the finish of 2024, we introduced Challenge SPARROW, an AI-powered edge computing resolution designed to autonomously acquire biodiversity knowledge in essentially the most distant corners of the planet. These units can run for lengthy intervals of time on solar energy with out disrupting the ecosystems during which they’re embedded, which is among the facets of the venture I’m most enthusiastic about. Defending important ecosystems advantages individuals, animals, and our planet. Our aim is to have SPARROW working on each continent by the tip of the 12 months.
In June 2025, I traveled to Nairobi to attach with our AI for Good Lab in Kenya and our companions on the bottom there. I was additionally invited to offer a TED speak centered on SPARROW and the alternatives it presents to extend biodiversity. I highlighted earlier the significance of partnerships with native organizations, and I view this as an opportunity to uplift our collaboration.
Utilizing AI to make new discoveries and enhance individuals’s lives
When researchers, conservationists, nonprofits, non-governmental organizations (NGOs), and educational establishments are tasked with doing extra with much less, expertise gives a approach ahead. AI transforms the way in which we will use knowledge to make new discoveries and enhance lives.
Microsoft is investing not solely in fixing at the moment’s challenges but additionally paving the way in which for a brighter tomorrow. Our work helps showcase the potential advantages of AI, pushing the bounds of what was beforehand doable. By making use of superior AI to real-world datasets, the AI for Good Lab helps unlock insights that empower communities and speed up progress on pressing international points.
Need to study extra? Listed below are a number of sources
- Discover the AI for Good Lab’s initiatives.
- Start your AI journey with AI sources and options from Microsoft.
- Take a look at our just lately introduced AI Economic system Institute, Microsoft’s company assume tank devoted to advancing unbiased analysis and creating actionable options that may assist societies worldwide positively adapt to the financial and social transformations introduced by AI.
