Artificial intelligence has been tasked with many projects in its brief history, but assisting in the latest United Nations projects aimed at ending world poverty, could represent the noblest goals yet posed to the artificial thinking of machine learning. It may also be the most challenging as its measurements and assessments would be applied to the entire world.
Ending world poverty by 2030 is the most ambitious of 17 global missions, planned by the United Nations in 2015. These Sustainable Development Goals will require substantial amounts of data and data analysis. Marshall Burke of Stanford University’s Earth System Science explained to Quartz the need for some measure to gauge the success of the project.
“How are we going to know if we’ve eliminated poverty if we don’t collect data? It’s like setting a weight loss goal but not having a scale to know if you’ve made any progress.”
Artificial intelligence would use machine learning to analyze satellite photos from all over the world, to pinpoint the exact location of people apparently living in very poor conditions. From this data, the team would create a worldwide poverty map that would be publicly available online. The Huffington Post reports Marshall Burke’s statements.
“We hope our data will be directly useful to governments around the world… to more effectively target their programs.”
Ending world poverty, to a great extent requires first finding the areas in the greatest need. Machine learning can be used to assist in locating the world’s neediest people and making that information available to organizations designed to help them. The information would be available to both governments and charities aimed at fighting poverty. In the past, resources like satellite imagery have been used for military intelligence, and purposes of war. Now that such images and the capacity for developing intelligence data are more commonplace, they can be used to help relieve human suffering.
Artificial intelligence would not only be able to spot the poorest areas and indicate the location of poverty-stricken villages, machine learning would be able to identify what exactly each village needed most. The AI uses a computer algorithm to identify signs of poverty through the process of machine learning.
Ending world poverty is dependent on discovering what people need, and where exactly these people can be found. The project will first aim at extreme poverty, defined as those living on less than $1.25 U.S. per day. Computers would look for predetermined criteria to identify and categorize the poverty photographed by satellites.
Artificial intelligence computers will analyze photos based on a variety of criteria including how much light emitted by a village at night. The computers will also consider the distance from roads, and waterways. It will examine urban areas as well as farmlands, searching for the needs of the poorest humans.
Ending world poverty in the past has been especially challenging due to the difficulty of obtaining survey information from remote villages. The project will be a vast improvement upon the survey methods that only managed to contact about 500 villages out of hundreds of thousands of such small communities. Only about half of the nations in sub-Saharan Africa have even conducted two or more surveys since 1990. Surveys are expensive, labor intensive, dangerous, time-consuming, reliant on human intelligence and are not very accurate. Remote locations are rarely reached at all.
With artificial intelligence analysis of data obtained by satellite images, need assessments will not depend on factors like distance, accessibility, preconceptions or bias. It will not be limited by any of the usual situations and inherent dangers that prevent those conducting surveys from reaching every village. It is now possible for the United Nations to be aware of nearly every human on the planet, assess their needs and make the information available for those who want to help.
Artificial intelligence and machine learning could make a huge difference in the timeline for ending world poverty.
[Photos by Getty Images/Getty Images]