Transcript: Episode 6

Welcome to Impact Signals, Social Impact at the Scale of AI. I'm Charlie. And I'm Sarah. Big day. The India AI Summit is in full swing, and some major announcements are reshaping how AI meets humanitarian work. Let's start with the World Food Program. They showed up to the India AI Summit, not with slides, with operational tools, grain ATMs, smart warehouses, route optimization, all running at scale inside India's public distribution system right now. That's what moves the needle, not pilot programs, production systems. India's PDS feeds hundreds of millions. The grain ATMs deliver precise food entitlements through fully digital systems. Smart warehouses use sensor-based AI to catch food losses before they happen. And route optimization cuts both costs and carbon emissions in one stroke. It needs $13 billion to assist 110 million people in 2026, and they're facing funding shortfalls that are forcing ration cuts. Exactly. AI can make every dollar go further, but it can't create dollars that don't exist. The efficiency gains matter most when resources are shrinking. And right now, they're shrinking fast across the humanitarian sector. Now, Google dropped some serious commitments at the summit. $60 million across two new challenges. $30 million for AI and government services, $30 million for AI and scientific research, plus a center for climate technology in India and a $15 billion infrastructure investment. The government challenge is fascinating. They cited a stat 74% of public servants globally already use AI, but only 18% believe their governments deploy it effectively. That gap between individual adoption and institutional effectiveness is exactly where the money needs to go. And Google DeepMind established a national partnership with India for frontier AI and science and education. That's not a grant. A long-term structural commitment. Right. The pattern is clear. The India summit is producing actual investments and partnerships, not just declarations, which matters because this is the first major AI summit hosted in the global south. And it's setting the tone for how development focused AI governance works. Speaking of the summit, they set a Guinness World Record, 250,000 responsible AI pledges in 24 hours, organized by India AI mission and Intel India. I'll be honest, records are symbolic, but the signal matters. The UN High Commissioner for Human Rights, Volker Turk, put it bluntly at the summit. AI has incredible force for good, but when you exercise power, you also need guardrails. A quarter million people agreeing to that in one day shows the demand for responsible frameworks isn't just a Western conversation anymore. Pivoting to disaster response, NASA published its analysis of cyclone gazani using satellite AI, MODIS, Landsat, image data, all deployed to map the damage in Madagascar. Category three cyclone, 200 per hour winds, at least 63 dead, 16,000 displaced, 25,000 homes destroyed. And this hit just 11 days after cyclone phidea. Madagascar is getting pummeled. The Copernicus Emergency Management Service activated rapid mapping, and the Southern African Development Community deployed a geospatial team alongside map action and the World Food Program. What stands out is the speed. Satellite AI goes from data capture to damage assessment in hours, not weeks. The ICRC also launched its appeal. The detail that got me was the missing persons figure in Ukraine, 89,000 people missing in action. The ICRC is using AI-assisted case management to process those cases and predictive modeling for anticipatory action. When you're dealing with that scale of human loss, AI isn't a nice to have. It's the only way to not fall impossibly behind. That's the opportunity in one frame. Food insecurity data meets national AI investment. Pakistan is one of the most climate vulnerable countries on Earth. If that billion dollars connects to early warning systems and precision agriculture informed by the IPC data, it could fundamentally change outcomes. If it doesn't connect, if it goes to chatbots and smart cities instead, then the food crisis just gets worse. And the multi-crisis watch is heavy today. Indonesia, severe floods, 2400 displaced in central Java. Sudan refugees still surging into chat. Sri Lanka issued a disaster situation report. Burkina Faso needs help for 2.7 million people. 21 forest fire alerts across three continents. Pre-Asian forecasting pools that can process all of it at once. That's episode six. Subscribe wherever you listen. Apple, Spotify, YouTube, and visit impactsignals.ai for the full research and sources. Stay ready.