First of all, this guy seems like he watched the last season of Silicon Valley and took it too seriously (perhaps due to having his firm name-dropped, who knows

First of all, this guy seems like he watched the last season of Silicon Valley and took it too seriously (perhaps due to having his firm name-dropped, who knows). > As Levine puts it, “Think about a self-driving car, it’s effectively a data center on wheels, and a drone is a data center with wings and a robot is a data center with arms and legs and a [ship] is a floating data center…” A self-driving car is not a datacentre. It will have a few computers, sure, but a datacentre is a distinct class of entity featuring hundreds of highly-available, enterprise grade computers; the self-driving cars that will ultimately succeed will have automotive grade machines, built to the lowest possible standard and price that meets quality and reliability requirements as defined by the DoT and the warranty department. A drone is not a datacentre with wings. It has the least it can possibly have aboard, because it needs to use as little electricity as possible for computing so that most of it can be used for spinning motors. While some drones do have multicore CPUs and GPUs on board, if anything we'll likely see improvements in wireless that enable that kind of tech to be pushed back to base stations / controllers / phones, so as to free up more precious battery for flying. Now, a ship I'll believe - you do need racks of servers on a cruise ship, for sure, especially given the fact that it's going to be away from a reliable network connection for ages at a time. However, it's not doing the same thing as the cloud; it's got to be reliable as hell with minimal IT staff on hand to deal with it, which means imaged machines running simple, basic things that can be trusted without having to have techs use expensive satellite data to fix it remotely. While there's some logic to the circular nature of centralized vs. distributed communication, packaging this up and branding it as "edge computing" strikes me less as a true observation or prediction, and more as someone who's spent too much time thinking about things through a marketing lens.