Gongol.com Archives: February 2016
Facebook can only really grow if the billions of people who don't have reliable Internet access become Internet users and join the site, so the company has a vested interest in expanding Internet access all over the world. In order to do that efficiently, they need to know where the people are. Thus the company is working on taking artificial intelligence and applying it to known data about the world (like satellite imagery) to come up with much more granular detail about where people can be found. They're having the Earth Institute at Columbia University review the data for quality, and Facebook then says it will make the data available on an open-source basis later this year. Facebook estimates that about 3 billion people worldwide have Internet access, and 4 billion don't. The population maps are mainly useful to Facebook when seeking to decide where to use wireless hotspots, where to use cellular-type service, and where they might have to turn to satellites or UAVs to deliver connectivity. It's estimated right now that 95% of the world's population is within reach of mobile phone service, but if those estimates are based on faulty data, then it may impede the necessary infrastructure investments to expand access. That's where better population-density mapping has a role to play. Of course, the research is being done with Facebook's private benefit in mind, but the spillover benefits from better mapping have the potential to do a lot of social good, like aiding in disaster planning and recovery.
The bookstore touts itself as a refuge from connectivity overload, but isn't the idea of a bookstore fundamentally to connect people with access to more information than they could possibly ever want to use? Maybe it makes people feel better, but disconnecting isn't necessarily a better way of life.
Google took billions of photos that included location data and fed them into a database. They then turned that database into a system that tries to identify the locations shown in new pictures based upon what it already knows about the rest of the world. Naturally, it's working better in places like tourist destinations that are well-documented than in remote areas, but it's apparently generally much better than human beings are at the same test. The Google system was able to at least get to the right continent about half of the time.
Mercedes is replacing robots in some of its plants with human workers, because it's easier to give a person detailed instructions than it is to reprogram the robots. Mercedes is trying to deliver more customized vehicles right off the assembly line, and people are their most efficient choice for now. This is actually a lesson learned long ago by Honda, which emphasizes the value of using people to do work because people can improve and innovate while automation cannot. There's a role for both, of course. We're better off when machines augment or supplement human work, labor, and thinking.
It's time to stop celebrating ignorance. As Ben Franklin said, "Being ignorant is not so much a shame, as being unwilling to learn."
Relitigating the Civil War may be one of the stupidest pursuits out there. Trump's only philosophical loyalty is to expediency, and that appears to be attracting some pretty shameful political bedfellows. It's also creating friction with our friendly neighbor, Mexico, where a former president has flatly rejected the idea that a Trump administration could somehow force Mexico to build and pay for a border wall.