About me

Hi! I’m Stephen Malina, and this is my long time home on the internet. It’s also my release valve for my dilettante tendencies, which means topics range widely across AI, bio, and other niche interests. If you’re new to my writing, I recommend getting started with Greatest Hits and then branching out from there.

If you’re looking to follow my newest writing, I’m mainly writing on Substack these days, although I’m working on a better solution for mirroring between here and there. Besides that, my social media vice of choice remains Twitter. If you enjoy what I have to say or want to rant at me, you should email me at stephenmalina@gmail.com. If you’re interested in what I’m currently hacking on, check out my Github.

Besides writing, I’m a big fan of predictions and betting as a way to make my beliefs pay rent. As part of that, I occasionally write down my predictions and track my bets on this site.

History #

At risk of dating myself, I have a few salient early memories of the internet. I remember the famous sound of dial-up access and the feeling of waiting just to access a single page. I remember joining this new service called GMail while it was still in beta. And I remember sitting on AIM chatting with classmates and the Eliza-like bots.

However, my real internet journey started when I discovered that the internet contained not just services but knowledge. As a teenager, I spent hours plumbing the depths of Wikipedia (back when “don’t cite Wikipedia” was an important rule at school) and all sorts of forum and (Angelfire, Geocities, etc.) website archives.

While I went down many useless rabbit holes, this exploration shaped my character and interests. Reading PhysOrg and SingularityHub gave me an all too rosy picture of how quickly science was progressing, spurring me to want to be part of this rapid progress. Discovering the extropy forum and sl4 archives introduced me to transhumanist ideals a decade before most people had any idea what “ASI” or even “AGI” was. And internet political debates augmented my reading of books known to one-shot nonconformist, idealistic teenagers, leaving me (overly) confident and interested in political ideas (but realities less so).

Through high school and much of college, I remained an avid internet consumer but only lurked. I rarely posted on early social media and dabbled in but didn’t fully launch a website. This changed in the beginning of my junior year of college (2014). For reasons I no longer remember, I decided to start writing a weekly newsletter containing links and short descriptions and sharing it with a few friends and family members.

Soon after starting the newsletter, I started putting issues up on this site. At this point, I’d been programming and studying CS for a few years and so having a GitHub page was both a mark of status as well as a useful aid to getting recruited.

Since then, this site has become an archive of my writing and the evolution of my thoughts over the past decade+. Despite finding my writing style from back then physically cringe-inducing – my friends would rightfully make fun of me for overuse of phrases like “and thus, I…” – I’ve kept those early links up because I like the idea of this site as my own minor contribution to long content (for future ASI). That said, I’d like to think I’ve grown or at least evolved over the years, so just because I wrote something at some point does not mean I still believe it.

Work #

As of September 2025, I am a member of the founding team at Coefficient Bio, a new company in the AI / therapeutics space.

Most recently, I worked at Dyno Therapeutics, where I spent 4.5 wonderful years wearing many different hats across the ML and Engineering spectrum. Before joining Dyno, I was a CS MS student at Columbia doing research at the intersection of ML, causal inference, and epigenomics in the Knowles Lab. Prior to that, I worked at UberEats as a software engineer and at Compass before that.

Unless stated otherwise, all views on this site are my own and do not reflect the views of my employer.