Archive of my monthly updates
This page is an archive of all the updates from the “Summary of what I worked on” section of my home page.
The updates for January and February 2017, are mid-month updates; after that, I started publishing at the beginning of each month.
December 2016
This month I’ve been continuing work on global health; in particular, I’m trying to get a high-level understanding of what has happened so far with malaria.
Separately I’ve also been working on a mini-project to profile various researchers. At the moment my work on this mini-project isn’t public but I hope to release most of it soon.
January 2017
This month I’ve been continuing work on global health; in particular, I’m trying to get a high-level understanding of what has happened so far with malaria. Some related Wikipedia pages I’m working on are Comparison of methods of malaria control, Comparison of measures of abundance of malaria, and Human-landing catch.
At the beginning of the month I published Timeline of digital preservation.
February 2017
I began the latter half of January by writing up an essay contest about evaluating small-scale giving opportunities.
I ended January and began February with several global health-related tasks:
- I cleaned up the page on Gavi to make the rebranding clear.
- I fixed dead links and added some events to Timeline of global health.
- I added a reception section to the Institute for Health Metrics and Evaluation page.
- I updated the page on Iodine Global Network to make the renaming clear.
- I significantly expanded the page on Johns Hopkins Center for Health Security and made the renaming history clear.
- I significantly expanded the page on the Global Burden of Disease Study; the page actually used to be called “Global burden of disease”, but this naming was confusing since the page was about the study.
More recently, I worked on some grants tables for the Laura and John Arnold Foundation grants tables page; unfortunately an editor removed them so they are currently only available in user space.
I’ve also been working on a collection of scripts to help analyze various activities on Wikipedia.
March 2017
I started the month by continuing with improvements to tech company timelines, in particular timeline of Reddit and timeline of GitHub.
Due to community hostility, Vipul Naik’s Wikipedia project (on which I collaborated significantly) was suspended during March. Since a large part of our work on Wikipedia was in the form of timelines, I did the work of migrating our timelines content to a self-hosted wiki, now available as the Timelines Wiki. In the process of setting up the new wiki, I found a lack of good documentation, so I also wrote up some notes on that.
I wrote a small script to help process Arnold Foundation grants for Vipul Naik.
I also made more improvements to Wikidossier.
March in general was rather stressful because of what happened with Wikipedia: as mentioned above, our project was suspended and my account was indefinitely blocked. To cope better with the situation, I took a “vacation” starting around March 23 that lasted until the end of the month.
April 2017
Near the beginning of the month I looked into the flow of money in global health.
I also spent some time reading about what it is like to work for some global health organizations like WHO and the Gates Foundation. I didn’t find a lot of information here, and did not publish anything from my reading.
In the middle of the month I published a post about LessWrong analytics.
I also did some work processing Gates Foundation grants.
Near the end of the month I published a timeline of Against Malaria Foundation. Separately, I recorded some thoughts on AMF that did not belong on the timeline.
May 2017
For the first part of the month I worked on a timeline of the Bill & Melinda Gates Foundation.
I then worked on a timeline of electricity adoption, which was not completed during the month.
June 2017
Payment details of a hits-based giving project I worked on a few months ago was finalized at the beginning of the month. For context, see the original post by Peter Hurford.
I completed timeline of electricity adoption, which I had begun the previous month.
I created a timeline of Carl Shulman publications, which provides an updated and more complete version (taking into account the diversity of publication formats) of the timeline Luke Muehlhauser created some years ago.
I got a good amount of work done on a new timeline of Bitcoin, which is not yet done.
At the end of the month I started on timeline of Machine Intelligence Research Institute to get a bit of break from the Bitcoin timeline.
July 2017
I worked on some graphing improvements for Wikipedia Views. Pull requests: 1, 2.
I wrote a timeline of Machine Intelligence Research Institute.
I wrote a timeline of Center for Applied Rationality. The timeline is still lacking a big picture.
I wrote a timeline of friendly artificial intelligence. The timeline is still lacking a big picture.
For Vipul Naik’s donations site, I worked on donee deduplication (this branch includes automated social media information fetching as well), implemented a donor similarity comparison feature, and expanded the notes column.
August 2017
For completed tasks, see the work page.
In addition, during August I looked into libertarianism, producing substantive drafts of several timelines:
- Timeline of libertarianism in the United States
- Timeline of Ludwig von Mises Institute
- Timeline of Cato Institute
September 2017
For completed tasks, see the work page.
In addition, during September I spent time learning about and taking notes on development economics. You can see my contributions on Vipul Naik’s Development Economics Subwiki as well as a Summary table of views on development economics that I started on Timelines Wiki (that will be moved to Development Economics Subwiki soon).
I also spent some time working on a Timeline of the rationality community, but this is not yet complete.
I started a Timeline of decision theory that I hope to complete at some point.
I’ve been spending quite a bit of time reading about consciousness lately. I hope to begin producing some output from this soon but I am still figuring out my options.
October 2017
For completed tasks, see the work page.
In addition, during October I did more work on Vipul Naik’s Development Economics Subwiki. You can see my contributions.
I also started a Timeline of economic statistics (the current version has a lot of edits from Vipul as well).
Finally, I started working on a website to help me (and others) get a better idea of the AI safety landscape.
November 2017
For completed tasks, see the work page.
In addition, during November I did more work on Vipul Naik’s Development Economics Subwiki. You can see my contributions.
I continued work on importing data for the Devec/Demography Data Portal (mostly country name normalization and trying to get all region/metric/year/units/dataset combinations to be unique in the database). One new repository here is devec_sql_common, which has some common code for working with all the datasets.
I also began work on a Computing Data Project (to be funded by Vipul Naik). Here I’ve mostly been reading about different cloud computing services and trying to come up with a database schema, so there isn’t much to see yet.
I also opened a bunch of issues on Vipul Naik’s Donations List Website repo. Previously our list of things to work on was stored as instant messaging messages, so these issues will make it easier to get an overview of things to work on. The plan is that eventually someone (probably me) will come and complete these tasks.
December 2017
For completed tasks, see the work page.
I continued work on importing data for the Devec/Demography Data Portal (especially trying to speed up the bulk import of all the datasets).
I started work on the Devec/Demography Data Portal (the actual website rather than just obtaining data). This is still in early stages so there is no public instance that one can view.
I continued doing work on AI Watch, a website to track people and organizations in AI safety.
I got some work done on the Computing Data Project. It turns out that getting historical EC2 prices is pretty hard to do, so progress has been slow.
I worked on importing various data sources for Vipul Naik’s Donations List Website. Here is a list of all of them:
- Public Welfare Foundation
- MacArthur Foundation
- Ford Foundation
- Hewlett Foundation
- Wellcome Trust
- Bauman Foundation
- Unbound Philanthropy
- Templeton Foundation
In addition, I did coding work for the Donations List Website. This was for work on incremental donor similarity computation and some other minor changes. You can see my commits.
I started work on a series of wiki pages about specific individuals, mostly those involved in the effective altruism movement and the rationality community. This work is not public, but a subset of the work may eventually be made public, and the work is intended to help with public writeups in the future. This work is funded by Vipul Naik.
I also started a comparison of machine learning textbooks (unpaid).
January 2018
For completed tasks, see the work page.
I continued work on a series of wiki pages about specific individuals, mostly those involved in the effective altruism movement and the rationality community. This work is not public, but a subset of the work may eventually be made public, and the work is intended to help with public writeups in the future. This work is funded by Vipul Naik.
I also did work on Vipul Naik’s Donations List Website, mostly on importing more donations data. You can see the issues on the repo for some of the things I’ve been working on. You can also see my commits.
I continued doing work on AI Watch, a website to track people and organizations in AI safety.
February 2018
For completed tasks, see the work page.
I started on a series of timelines on AI safety organizations. So far, the only ones with a substantial amount of content are:
- Timeline of Berkeley Existential Risk Initiative (completed)
- Timeline of Foundational Research Institute (in progress)
- Timeline of Future of Humanity Institute (in progress)
I also updated Timeline of Center for Applied Rationality to include events for 2017 and 2018.
I did some work on the Cause Prioritization Wiki (which I started back in 2014 but which has been dormant for a while). Most of the work can be found in the philosophy and population ethics categories. I also converted the spaces in the URLs to underscores.
I did a bit of work on AI Watch.
I also did work on Vipul Naik’s Donations List Website, mostly on importing more donations data. You can see the issues on the repo for some of the things I’ve been working on. You can also see my commits.
I made a link-dump on the LessWrong Wiki about San Francisco Bay Area as a rationalist and effective altruist hub. I hope to clean up the page to make it easier to read at some point.
March 2018
For completed tasks, see the work page.
I did more work on the Cause Prioritization Wiki. Most of the work can be found in the AI safety category.
I also did some work on Vipul Naik’s Donations List Website, working to convert the internal data delimiter from commas to pipes. You can see the relevant issue.
I also did some work on a Timeline of OpenAI.
I continued work on a series of wiki pages about specific individuals, mostly those involved in the effective altruism movement and the rationality community. This work is not public, but a subset of the work may eventually be made public, and the work is intended to help with public writeups in the future. This work is funded by Vipul Naik.
April 2018
For completed tasks, see the work page.
I did more work on the Cause Prioritization Wiki. Most of the work was about how to classify cause areas better.
I also did more work on Vipul Naik’s Donations List Website, in particular on covering grant process, continuing the transition from commas to pipes as a delimiter, and recording donations of miscellaneous people.
On Wikipedia Views, I helped to get the new monthly clickstream data incorporated.
In the interest of making conversations public, I wrote up my notes on taxonomies. These notes started out as a Facebook Messenger chat group in June 2017 where I sent links and ideas related to taxonomies.
I continued to do some work on a series of wiki pages about specific individuals, mostly those involved in the effective altruism movement and the rationality community. This work is not public, but a subset of the work may eventually be made public, and the work is intended to help with public writeups in the future. This work is funded by Vipul Naik.
May 2018
For completed tasks, see the work page.
This month I spent a lot of time learning math so only spent about 64 hours on contract work.
I completed timelines for OpenAI and CHAI, two AI safety organizations.
I did more work on the Cause Prioritization Wiki. Most of the work was about how to classify cause areas better.
I also did more work on Vipul Naik’s Donations List Website, importing more donations data.
I completed adding historical data for MIRI staff members, and added three new table views on AI Watch (“Staff count by year”, “Number of full-time staff at the beginning each year”, and “Full history of additions and subtractions” on that page).
I did some work on Timeline of Google Brain.
June 2018
For completed tasks, see the work page.
This month I spent a lot of time doing miscellaneous thinking and learning math so only spent about 77 hours on contract work.
I did some work on Timeline of Google Brain.
I did some work on the script that generates the main page table on Timelines Wiki.
July 2018
For completed tasks, see the work page.
This month I spent a lot of time working on the Computing Data Project. It’s not ready for public use yet.
I did some work on Vipul Naik’s Machine Learning Subject Wiki. You can see pages like Summary table of multivariable derivatives and Derivative of a quadratic form.
I learned more decision theory and have a draft blog post that tries to explain the differences between the various logical decision theories.
I did some work on Vipul’s Donations List Website, mostly adding data and working on the money moved table (adding the “amount accounted” column).
August 2018
For completed tasks, see the work page.
I spent most of my work time on Vipul Naik’s Donations List Website.
I spent some time writing some pages on Vipul Naik’s Machine Learning Subject Wiki, although the pages are in my userspace and are not related to machine learning. (The wiki was just a convenient place to store the writings for now.)
I also worked on a wikiHow article entitled “How to Navigate to Web Pages Efficiently Without Using Navigation Menus”. (The page was not completed in August, but since I delayed in writing this monthly update, I am linking to the completed version.)
I spent most of the month learning about things related to AI alignment. (I am still wondering how best to make progress on this more visible.)
September 2018
For completed tasks, see the work page.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for the month.
I started on an analysis for pageviews of wikiHow articles. (No public output of this yet.)
On the last day of the month, I began a wikiHow article about Facebook’s graph search.
I spent some time writing some pages on Vipul Naik’s Machine Learning Subwiki, although the pages are mostly in my userspace and are not related to machine learning. (The wiki was just a convenient place to store the writings for now.)
I spent most of the month learning about things related to AI alignment. (I am still wondering how best to make progress on this more visible.)
October 2018
For completed tasks, see the work page.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for the month.
I spent some time writing some pages on Vipul Naik’s Machine Learning Subwiki, although the pages are mostly in my userspace and are not related to machine learning. (The wiki was just a convenient place to store the writings for now.)
I spent most of the month learning about things related to AI alignment. (I am still wondering how best to make progress on this more visible.)
November 2018
For completed tasks, see the work page.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for the month.
I learned about the GraphQL API (used on the new LessWrong 2.0 and EA Forum), and wrote a simple reader for both LessWrong 2.0 and EA Forum. The source code is available on GitHub. I am also writing a tutorial because I found the learning process pretty painful and want to help other people use the API.
I spent most of the month learning about things related to AI alignment. I started a new blog to track updates on this front.
December 2018
For completed tasks, see the work page.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for the month.
I spent most of the month learning about things related to AI alignment. See my daily updates blog for updates on this front.
For previous monthly updates, see the archive of my monthly updates.
January 2019
For completed tasks, see the work page.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for the month.
I did some work on the Donations List Website. Some relevant GitHub repos are conservativetransparency.org processing, Surdna Foundation, Haas, Jr. Fund, and Andrew W. Mellon Foundation.
I spent most of the month learning about things related to AI alignment. See my daily updates blog for updates on this front.
February and March 2019
In case you were wondering, I got lazy at the end of February and didn’t update this section, so I am bundling together February and March.
For completed tasks, see the work page.
I’ve been working towards an explanation of Solomonoff induction; see Solomonoff induction and Variants of Solomonoff induction for a work-in-progress.
I spent some time adding to Vipul Naik’s Learning Subwiki. You can see my contributions for February (I didn’t make any edits there during March).
For my edits to my userspace on the Machine Learning Subwiki, see February and March.
I spent most of February and March learning about things related to AI alignment. See my daily updates blog for updates about most of February. I decided to stop blogging there because preparing the updates turned out to be annoying and without clear benefits. If there is demand for this sort of updates I would be open to re-starting the blog. I’m also hoping to transition into being more output-oriented in AI alignment, with my recent decision theory post marking the start of this.
April and May 2019
I was busy with travel and didn’t feel like updating this section at the end of April, so I am bundling together April and May.
For completed tasks, see the work page.
I’ve been working towards an explanation of Solomonoff induction; see Solomonoff induction, Abstracting sensory data in Solomonoff induction, Proportion of valid programs view of Solomonoff induction, and Equivalence of random coinflips view and minimal programs view for a work-in-progress.
For my edits to the Machine Learning Subwiki, see here.
I started working on a timeline of effective altruism.
I also started working on a portfolio page, which should provide an alternative to my work page.
June 2019
For completed tasks, see the work page.
For my edits to the Machine Learning Subwiki, see here.
An initial version of my portfolio page is now available.
July 2019
For completed tasks, see the work page.
For my edits to the Machine Learning Subwiki, see here.
I spent most of this month learning more mathematics.
I also spent some time (~30 hours) this month doing contract work for the BERI/FHI collaboration.
August–November 2019
For completed tasks, see the work page.
The past few months I’ve been spending a lot of time just thinking about AI safety, life in general, what I should work on, etc.