I'm still one of those rare weirdos who use RSS feeds (with Feedly) and over the years I have started following many people and reading many blogs. Unfortunately, many of them have stopped posting, but there are still some people who regularly post really interesting stuff. I am impressed by how some of them can sustain posting daily, even if only a short snippet, others post 2-3 times a week some interesting observations in a wide range of fields, and others keep posting coding optimizations even though they are tenured professors, but they are not afraid to get their hands dirty.
I hope I will soon be able to post more often and to post stuff that is more interesting and have a wider audience again. In the mean time, I'll start sharing with you some of the blogs I'm following, with a short description:
John D. Cook: This guy is a statistician by training, who does consulting work and he posts on all kinds of stuff, from numerical computation to solving real business needs. One of my favorite posts from him is Random is as random does, which I find is a useful reminder whenever you read an article about something that is modeled probabilistically, that just because that model fits, it doesn't mean that the model is the underlying reality. I stumbled upon John on StumbleUpon (wow, I feel old now).
FastML: This a machine learning blog ran by several people under one pseudonym. They have a funny writing style, like to criticize many novelty papers, but most importantly, they give lots of useful code that can be directly used, such as tricks on how to encode capital/lower letters for RNNs. Really cool stuff.