First research update!

[Ram Rachum]

Hello, everyone!

You're getting this email because you're one of the 114 Googlers who signed up to get updates on my research into using machine learning to understand our society.

Four days ago, on March 1st, I turned in my gun laptop and my badge, and officially became a free agent again. Google was an amazing place to work and I'm sad to leave, but I'm excited about working full-time on my research.

You've probably watched the talk I gave inside Google. (Slides.) I gave 7 runs of that talk to different Google offices around the world. In total, 2,300 Googlers viewed my talk live. Many others viewed the recording at go/ramrachum-research-talk. I've gotten around 50 chats and emails from people who are excited about the research, and shared relevant research and books with me. I've made a reading list for myself based on that. It might take me a year just to go through all of them!

One benefit of this group is that hopefully some of you could help me in areas I'm not very experienced in. I've included two such calls-to-action in the section below. It's possible that in the future I'll also need a mentor in academic writing, but that's at least 6 months away.

Goals for March

I think it'll be cool if I'll send an update to this list at the start of every month. On every update, I'll share my goals for the month and the progress I've done with the previous month's goals. I'm not sure that this format will hold up, but let's give it a try. 

1. Make my talk available on YouTube: You've all seen my talk inside Google, but that recording can't be shared with outsiders. I now made a public recording of my talk and posted it on YouTube! This is very important for me, because now I can send that talk to people I want to work with, and have them understand what I'm about. This is especially important so I could...

2. Try to get a grant. As I said in the talk, I'm planning on spending the next 6-12 months of my life working on the research, burning through my savings. This is a huge risk that I'm taking. If I could get a grant of some kind, that would be awesome. Fortunately, pinkgothic@ connected me to Scott Alexander, who recommended the Long-Term Future Fund, the Survival And Flourishing Fund and Open Philantropy. I'm not holding my breath on any of these, but if I could get a grant that would be very helpful. I'll send them the YouTube video, along with lots of information on my research goals. Do you know of any other grant programs that would consider funding an independent researcher like me? If so, please connect me!

3. Set up research infrastructure: One of my challenges in my new life as a researcher would be to organize all the information I've got, such as research papers I was sent, in a way that's easy to work with. I'm good at this kind of organization, but my younger brother is better. He taught me how to use Obsidian, which is sort of a... WYSIWYG markdown editor that lets you build your own private Wikipedia on local markdown files. I love this. He got the recommendation for that program from CGP Grey, who we both idolize. (Example video, anyone who's interested in explaining science to the public should know Grey's videos.)

4. Take a vacation and clear my mind. I went through lots of difficult changes these last 3 months. I really deserve a vacation. Somewhere near the end of March I'll probably fly to London, and from there probably to New York. I haven't set the dates yet, but I'm thinking that 3 weeks would be a good length. I haven't been outside of Israel for 2 years, and I miss that so much!

Here's a goal that I'll likely postpone to April: Improve the UI for the Marley framework. I haven't talked too much about the Marley framework yet, even though it's at the center of my work. This is the framework I'm working on that will let me iterate on experiments faster. I'll talk about it more when there's more interesting things to show. Right now, I need to solve a few architectural problems in the UI, which is written in Vue.js. I'm a beginner at Vue, and you can see some of my mediocre Vue code here. Are you an expert on Vue, and will you have time to mentor me in April? If so please contact me.

Have a good month,

Ram.