Over the decades, many hundreds of millions across the world have benefitted from charity. I daresay each and every one of you has made at least one donation in your life, be it to support the fight against global issues––the refugee crisis, climate change, natural disasters––or even to help fuel the annual Clash of Cans event at PCMS. But people who need help the most sometimes miss out on it. This isn’t intentional by any means, of course; it’s not the result of underlying discrimination or bias or anything malicious of that sort. Ignorance is most likely the culprit. We simply aren’t aware of who needs what, or who needs what when, and in terms of charity, it’s a major oversight.
But the problem has largely been solved by GiveDirectly, a charity organization that makes donations of cash money as an alternative to food or blankets, because “cash enables choice.” They’ve been at it for the past decade, and they have helped the impoverished all across the world. Very recently, GiveDirectly partnered with UC Berkeley to design an algorithm that can calculate which people in a particular area in a particular region needs the most help. It accomplishes this in two steps: First, the algorithm uses high-quality satellite images to determine which regions of the world are the most impoverished, or are most in need of help. Then, once the target region has been identified and the search has been narrowed considerably, the algorithm analyzes mobile phone data to distinguish between the rich and the poor.
So far, the algorithm has had much success. The project was launched in Togo, a country in West Africa, in which some 55% of the population subsist on less than $1.90 a day. With the algorithm’s help, GiveDirectly plans to donate a total of $5 million dollars.
It’s nice to hear of yet another instance of an algorithm lending a helping hand.