photo illustration -Human intelligence provides the purpose, judgment and responsibility. Artificial intelligence expands what humanity can achieve.
In Nigeria, an early forecast became something families could use: money to move children, protect food and prepare before the water peaked.
By: Ya Libnan – Third article in a series about how AI is improving our lives : Analysis
Imagine being told that floodwater may soon enter your home. You know you should move your children to higher ground, buy food and medicine, and protect the harvest you depend on. But what if you cannot afford to do any of those things?
That is why a promising use of artificial intelligence deserves our attention. AI can help predict a flood. When people turn that prediction into timely assistance, families gain something precious: a chance to act.
In 2024, the humanitarian organization GiveDirectly worked in Nigeria’s Kogi State, along the Niger River. Its team used flood forecasts, including data from Google’s AI-powered Flood Hub, together with local knowledge to decide when to send cash to people at risk. Residents helped identify flood patterns and monitor water levels. The computers supplied an early signal; people helped decide how to respond.
GiveDirectly reports that it sent early payments to 3,250 people between August 29 and October 6, before the floods peaked on October 21. The first transfer was $105. Families used the money to move children and older relatives to safer ground, buy essentials, and build bamboo rafts or raised platforms to protect food and belongings. One recipient told the organization that the transfer helped her move property that might otherwise have been damaged.
Those are modest actions when measured against the force of a flood. To a family facing that flood, they can mean everything.
The story also shows the limits of a forecast. AI did not stop the water, and the flood still damaged homes and farms. GiveDirectly reported that some payments were delayed while it resolved identity checks. People without phones could also be harder to reach. These details matter because a warning helps only when it reaches the right person in time and comes with a practical way to respond.
The result was a partnership among researchers, aid workers and local residents. Google’s models helped identify where and when danger might arrive. GiveDirectly arranged the transfers. Community members brought knowledge that a model alone could miss. Families made the decisions that mattered most for their own safety.
This approach could be especially valuable where rivers are poorly monitored and conventional warning systems are limited. Google says its Flood Hub now offers forecasts for significant river floods in areas covering about two billion people in more than 150 countries. Coverage is an opportunity, not proof that every resident receives or can act on a warning. Governments and aid organizations still need trusted local networks, reliable delivery systems and plans for those who are easiest to overlook.
We began this series to show how AI is already improving lives. The families in Kogi give us a clear example. The most meaningful measure of a technology is not how impressive its prediction looks on a screen. It is whether a mother has time to move her children, whether a farmer can save food, and whether a family can face the next morning with more than they would otherwise have had.
AI gave an earlier warning. Human beings turned it into help. That is a success story worth telling—and a model worth improving.
