Starting January 15, 2014, I did contract work for the Machine Intelligence Research Institute (MIRI). I worked under Luke Muehlhauser, who was the director of MIRI at the time. I was recommended for the position by Jonah Sinick, and interviewed with Luke in November 2013 via Skype. My work for MIRI fulfilled part of my responsibilities under Optional Practical Training (OPT) (see my immigration status page for more).

Below, I list and link to the projects I’ve done for them that are publicly available.

• Distribution of computation in the world (quick and dirty overview): My work culminated in a blog post by Luke Muehlhauser titled The world’s distribution of computation (initial findings), published February 28, 2014. You can also access my notes for the project here.
• Forecasting and related topics: Most of my recent LessWrong posts relate to my work for the project. My concluding post on LessWrong is here.

I stopped work for MIRI in August 2014 prior to starting my full-time job at LiftIgniter.

You can also read my answer (and other answers) to the Quora question What is it like to do contract work for MIRI?

Luke Muehlhauser stepped down as director of MIRI in May 2015 and was succeeded by Nate Soares. I believe the work I did with Luke would no longer be sponsored by MIRI as their research agenda is now much more narrowly focused on the mathematical parts. However, my friend Katja Grace is pursuing work of a loosely similar nature at AI Impacts, which has received some of its funding through MIRI (these are funds specifically earmarked for AI Impacts, and not general MIRI funds).