The Tail at Scale. Great article about strategies to reduce latency and increase reliability in distributed systems.
Lambda Architecture. I may have posted this before, but I keep coming back to it.
Using Edgeworth series to approximate random inequalities. Super cool.
Good critique of append-only databases
Bigger Data; same laptop. Quote: "I’d like to challenge the teams behind the existing graph processing platforms to evaluate and report their COST (Configuration that Outperforms a Single Thread) for this dataset."
Column Oriented Database - Introduction, and part 2.
A nice tutorial on using Akka Streaming. See also this example application and the cookbook. I'm getting excited by this project - I was bullish on Scalaz Stream, but the performance leaves something to be desired.
Sketching and Embedding are Equivalent for Norms. This is a very nice result.
Augmenting Human Intelligence: A Conceptual Framework. This paper is a classic, by the man who invented the computer mouse, but I never got around to reading it until recently.
Build on Sand. Great article about how Singapore is importing sand to build new territory. Personally, I'm also very interested in simply building on the water - floating developments, ala seasteading.
Scott Alexander debunks some silly claims. Specifically, a recent article claims that perceptions of high innate ability requirements is correlated with fewer women in tech. Scott Alexander shows that actual high innate ability requirements is more strongly correlated, so perceptions are probably tracking reality. See also Scott's comment on why IQ deniers are dangerous.
Somehow I also missed Scott Alexanders explanation of the neoreactionary position when it first came out. I strongly recommend it - it's far more comprehensible (and shorter) than Moldbug or any real neoreactionary. The article is not a straw man - according to Michael Anissimov, it's close enough. I don't endorse the neoreactionary view, but I think anyone who dismisses it without thought is being foolish.
The Kafkatrap - when the act of demanding a definition of X that can be checked and falsified proves you are X.
I've been watching a reasonable amount of TV lately. I strongly recommend watching Archer and The Boondocks. Weird observation: to me, The Boondocks seems chock full of references that would only make sense to Americans, but somehow every African I know loves it.
Over the past 15-30k years, dogs and wolves diverged evolutionarily. The net result is significant cognitive differences - dogs now interact with humans socially and solve problems using the humans as tools.
Notes on Boundaries - an excellent article describing how forming boundaries is important. This line of thinking is one of the strongest arguments against immigration which I'm aware of. But a strange fact is that most arguments against immigration, including this one, could be also be applied in favor of (racial, in-country) segregation, ala American Jim Crow. Strangely, I've never heard an immigration opponent actually state that if blacks crossed that cutoff, they'd favor segregation. This suggests that those arguments are not their true rejection. This article by Nyan Sandwich is interesting because it's one of the few such arguments which actually hints in that direction.
This is a very strange article on how easy it is to get into technology. The article discusses how an underprivileged kid (a non-Asian minority, if you care about such things) needed to (a) attend some free events and (b) get internet at home. Along the way, the kid receives a a bunch of free stuff. When he's ready to look for a job, he needs to put a code portfolio on a free website. The mood affiliation of the article is odd.