Facebook uses artificial intelligence to create the world's most detailed population map
To complete this project, Mark Zuckerberg's company has decided to capture high- resolution satellite images and use a machine that maps and takes photos of all houses on the planet.
Facebook wants us all to be connected but they need to know where people really are to be able to do this. However, since current population maps are not accurate enough for their needs, the team at Facebook have decided to create a map themselves.
Jael Maguire, director at Facebook's connectivity lab, Yael Maguire, director at Facebook's connectivity lab, has said that the aim of the project is to determine whether it is possible to develop technologies that allows us all to be connected. The best existing source of data about population is the map by the University of Columbia but this is not detailed since it only offers 1-km resolution. And if Facebook wants us all to be interconnected, the information in this map is a bit misleading.
Last year, Facebook's connectivity lab team met with the company's artificial intelligence group and a scientific team specialized in data to develop the first prototype/execution. They examined images from 20 countries totaling 21.6 million square kilometers. Facebook's artificial intelligence team processed and examined 14.6 billion images.
How did we do it?
To find evidence of human settlements in these images, the team used a pattern of learning techniques to prepare the algorithm that determines which images contain evidence of human settlements. It was also possible to use the machines in Facebook's data centers.
The result of their hard work is the world's most detailed population density map. Now Facebook would like to know the community's opinion on how this data could help to analyze connectivity. However, Mark Zuckerberg's company is currently focusing on connectivity, i.e. having this data used in other areas.