As we wrote about in September of last year, the supercomputing world was being dominated by Chinese entrants. For the first time in 5 years, the U.S. no longer sported a top 3 fastest supercomputer. The Titan supercomputer at Oak Ridge National Laboratory clocked a measly 17.6 petaflops, running up against the top two Chinese systems, clocking in at 93 and 33.8 petaflops respectively. And those differences aren’t just numerical bragging rights — they represent the ability to blaze the trail on AI,

Nearly every business leader (99% according to an HBR study) recognizes the need to build a data driven organization. However, so many struggle to use data for a variety of reasons, a big one of them being that data exists in all sorts of places. This is especially the case with startups, who use a number of out of the box tools that don’t connect to each other, and that only offer so much granularity and flexibility. The solution? A data warehouse.

A great article by Matt Gershoff that covers risk, reward, opportunity costs, and just generally interesting thoughts on experimentation. Basically, "if you just want to pick between A and B, do you really need to run standard significance tests at a 90% or 95% confidence levels?" Short answer, no. The article covers the logic as well as math behind why (and when) to just "go-with-it" (maybe even flipping a coin). Why is this article (and approach) helpful? Because it can help with what type of approach to take based on the problem.