1. It Begins
“Think where man’s glory most begins and ends,/ And say my glory was I had such friends"
Hello!
Every two weeks, I will send this short email that summarizes the previous two weeks, outlines a plan, and adds relevant reading links.
I’m trying to resist tendencies to neatly edit everything before sending. The aim is for it to feel like a natural compilation of everything I’ve done in the last couple weeks, rather than a finished product.
Writing
Essays
Why I’m Researching Open Source ML: because it’s fun and useful.
Questions about (Economic) Growth: one of a two part series about growth and governance.
Notes
What is a good taxonomy of ML tooling? (data preparation > data loading, experiment tracking and management, model specification, general infrastructure is current working model)
External Links
The New American University by Nadia Eghbal. An excellent long form post exploring Arizona State University’s unusual approach to funding higher education. The biggest takeaway for me was the possibility of decoupling prestige from exclusivity. We members of society are so conditioned to thinking that exclusive := excellent. I’ve been skeptical about this narrative (especially in the context of higher ed) but until now unable to articulate an alternative vision.
Algorithms as culture: Some tactics for the ethnography of algorithmic systems by Nick Seaver. Seaver’s main point is that we should embrace the nebulosity of the term ‘algorithms’ by treating them not as lines of code or specific procedures, but rather vague objects that are ‘enacted’ by different participants. In other words, the Spotify recommendation algorithm is a database of code, as enacted by engineers, but also the changing behavior of users in response to the recommendation, as enacted by users. Useful and entertaining, but not fully clear to me. [H/t Ziv]
Public Faith by Nadia Eghbal. Short post about the noticeable presence of religion among those embedded in the open source community.
Planning
Review
(this section will reflect on previous two weeks’ plan)
Next Two Weeks
Finish and write up user survey of most typically used/impactful ML tools
Get descriptive statistics and do simple analysis of the repos
(Stretch) write up short essay with results
Fill out form for MIT IRB