Michael Mullarkey
@mcmullarkey.bsky.social
Clinical psych PhD | ML Engineer in sports analytics| he/him
My newsletter on building cool data stuff: https://buttondown.com/datadash
Personal website: mcmullarkey.github.io
Opinions are my own and don't reflect the opinions of my employer
My newsletter on building cool data stuff: https://buttondown.com/datadash
Personal website: mcmullarkey.github.io
Opinions are my own and don't reflect the opinions of my employer
Pinned
Is social media to blame for the teen mental health crisis? Listen to me, a middle-aged libertarian, tell Joe Rogan that it is.
podcasts.apple.com/us/podcast/t...
podcasts.apple.com/us/podcast/t...
If Books Could Kill: The Anxious Generation on Apple Podcasts
Show If Books Could Kill, Ep The Anxious Generation - Aug 8, 2024
podcasts.apple.com
Honored to play a small part in making this episode!
Thanks to @michaelhobbes.bsky.social for reaching out to me. My job was mostly starting to write down "he should also talk about this study" and then said study getting mentioned before I finished writing
Thanks to @michaelhobbes.bsky.social for reaching out to me. My job was mostly starting to write down "he should also talk about this study" and then said study getting mentioned before I finished writing
A bitter irony: most screeners for problematic video gaming pick up on problematic *gambling*
First post of the year, new paper out today: we present possibly the biggest case of systematic Measurement Schmeasurement in tech use. It seems that most studies on gaming (videogame) addiction/disorder haven't measured gaming after all. This research took years, so long 🧵 doi.org/10.1098/rsos...
Confusion in gaming disorder measurement
Abstract. Measurement is important for the scientific programmes of addictive behaviours. In the present study, we investigated the measurement of gaming d
doi.org
January 28, 2026 at 5:30 PM
A bitter irony: most screeners for problematic video gaming pick up on problematic *gambling*
Reposted by Michael Mullarkey
I love this idea: duckdb.org/community_ex.... It translates dplyr syntax _inside_ of duckdb so you can mix it with regular SQL
dplyr
DuckDB Community Extensions R dplyr pipeline syntax support for DuckDB - transpiles dplyr verbs to SQL
duckdb.org
January 27, 2026 at 7:08 PM
I love this idea: duckdb.org/community_ex.... It translates dplyr syntax _inside_ of duckdb so you can mix it with regular SQL
There are lots of ways to help, and if you're as lucky as me a lot of organizations could use your $$
Here's a list of vetted resources in Minnesota passed along by a friend
Here's a list of vetted resources in Minnesota passed along by a friend
Stand With Minnesota Donation Directory
Stand With Minnesota is a hub for supporting, learning, and taking action to support Minnesotans impacted by ICE and federal enforcement.
www.standwithminnesota.com
January 25, 2026 at 8:03 PM
There are lots of ways to help, and if you're as lucky as me a lot of organizations could use your $$
Here's a list of vetted resources in Minnesota passed along by a friend
Here's a list of vetted resources in Minnesota passed along by a friend
Reposted by Michael Mullarkey
I am trying not to post on these events in this state of mind but: I hope people understand what the observers are doing is brave and dangerous.
January 24, 2026 at 4:51 PM
I am trying not to post on these events in this state of mind but: I hope people understand what the observers are doing is brave and dangerous.
Reposted by Michael Mullarkey
Introducing HiRO-ACE: an AI framework that makes highly detailed climate simulations dramatically more accessible. It generates decades of high-resolution precipitation data for any region in a day on a single GPU—no supercomputing cluster required. 🧵
January 21, 2026 at 7:34 PM
Introducing HiRO-ACE: an AI framework that makes highly detailed climate simulations dramatically more accessible. It generates decades of high-resolution precipitation data for any region in a day on a single GPU—no supercomputing cluster required. 🧵
Reposted by Michael Mullarkey
Also I think a bunch about how the lone genius frame of software convinces us this is the only way to succeed, so everyone is jamming their brain around trying to do the impossible in isolation, when a social frame means we can actually scaffold what each other knows
Encoding all information in the world into your memory is neither possible nor adaptive even if you could; it should not be something you pressure yourself to do at every moment
Attention directs memory formation for a reason
Attention directs memory formation for a reason
January 21, 2026 at 6:54 PM
Also I think a bunch about how the lone genius frame of software convinces us this is the only way to succeed, so everyone is jamming their brain around trying to do the impossible in isolation, when a social frame means we can actually scaffold what each other knows
Reposted by Michael Mullarkey
Hacking on an embeddable Python REPL, powered by Wasm/Pyodide.
My goal is to make it super simple to get Python running on a website...so that there's no package installation required, just specify the script tag and div and you're done (supports theming and package installs via attribute as well)!
My goal is to make it super simple to get Python running on a website...so that there's no package installation required, just specify the script tag and div and you're done (supports theming and package installs via attribute as well)!
January 21, 2026 at 5:04 AM
Hacking on an embeddable Python REPL, powered by Wasm/Pyodide.
My goal is to make it super simple to get Python running on a website...so that there's no package installation required, just specify the script tag and div and you're done (supports theming and package installs via attribute as well)!
My goal is to make it super simple to get Python running on a website...so that there's no package installation required, just specify the script tag and div and you're done (supports theming and package installs via attribute as well)!
Appreciate the effort to go beyond "median query" in estimating LLM energy usage for individuals
Would love to know where "several Claude Code instances using largest model for several hours per day" sits in the individual usage distribution. My gut-check guess is top 1% but we need better data
Would love to know where "several Claude Code instances using largest model for several hours per day" sits in the individual usage distribution. My gut-check guess is top 1% but we need better data
Whenever I read discourse on AI energy/water use that focuses on the "median query," I can't help but feel misled. Coding agents like Claude Code send hundreds of longer-than-median queries every session, and I run dozens of sessions a day.
On my blog: www.simonpcouch.com/blog/2026-01...
On my blog: www.simonpcouch.com/blog/2026-01...
January 21, 2026 at 2:54 AM
Appreciate the effort to go beyond "median query" in estimating LLM energy usage for individuals
Would love to know where "several Claude Code instances using largest model for several hours per day" sits in the individual usage distribution. My gut-check guess is top 1% but we need better data
Would love to know where "several Claude Code instances using largest model for several hours per day" sits in the individual usage distribution. My gut-check guess is top 1% but we need better data
Reposted by Michael Mullarkey
I started reading up on the whole "loneliness pandemic" narrative because this seems like a literature where the age-period-cohort problem may be relevant (or maybe it isn't?).
Here's data from Australia (HILDA), average agreement with the statement "I often feel very lonely" (SD of ca. 1.8).>
Here's data from Australia (HILDA), average agreement with the statement "I often feel very lonely" (SD of ca. 1.8).>
January 15, 2026 at 1:24 PM
I started reading up on the whole "loneliness pandemic" narrative because this seems like a literature where the age-period-cohort problem may be relevant (or maybe it isn't?).
Here's data from Australia (HILDA), average agreement with the statement "I often feel very lonely" (SD of ca. 1.8).>
Here's data from Australia (HILDA), average agreement with the statement "I often feel very lonely" (SD of ca. 1.8).>
"Sure he's a lot but he's good on technical stuff" is my sleeper activation phrase to run as fast as possible in the other direction
January 19, 2026 at 1:45 AM
"Sure he's a lot but he's good on technical stuff" is my sleeper activation phrase to run as fast as possible in the other direction
Is the mother of invention necessity or "I want this very specific weird/wonderful thing"
in my madness I'm designing a weird L-shaped partial enclosure for a 500w-48v power supply because I don't want a giant flat box on my bench
January 15, 2026 at 5:56 AM
Is the mother of invention necessity or "I want this very specific weird/wonderful thing"
Reposted by Michael Mullarkey
Magnificent. Make sure you click on the "1 more reply" link to make sure you read the whole thread. You will not regret it.
It's the 3rd of January 1939, your name is professor James Smith and you are staring at a drawing of a fish. It's not a regular fish. It is a completely impossible fish.
January 10, 2026 at 10:50 AM
Magnificent. Make sure you click on the "1 more reply" link to make sure you read the whole thread. You will not regret it.
If you're also at the end of your rope emotionally, I offer this live remix of No Broke Boys featuring a guitar solo
No Broke Boys 🤑‼️ Disco Lines, Tinashe: No Broke Boys (live remix) I know I’m soooo late to this one but I had to make a remix at some point 🥹 #dj #guitar #remix #fyp @discolines @Tinashe
TikTok video by Jarrod Costello
www.tiktok.com
January 9, 2026 at 7:51 PM
If you're also at the end of your rope emotionally, I offer this live remix of No Broke Boys featuring a guitar solo
I know I'm extremely late and it's unreal how much more I enjoy the For You feed vs Discover
January 9, 2026 at 5:30 AM
I know I'm extremely late and it's unreal how much more I enjoy the For You feed vs Discover
Reposted by Michael Mullarkey
An important tip to improve your 2026
January 9, 2026 at 1:58 AM
An important tip to improve your 2026
Reposted by Michael Mullarkey
I wrote a bluesky client from scratch in HTML and JS!
You can try it here: impro.social
github.com/improsocial/...
You can try it here: impro.social
github.com/improsocial/...
GitHub - improsocial/impro: A Bluesky web client
A Bluesky web client. Contribute to improsocial/impro development by creating an account on GitHub.
github.com
January 9, 2026 at 12:27 AM
I wrote a bluesky client from scratch in HTML and JS!
You can try it here: impro.social
github.com/improsocial/...
You can try it here: impro.social
github.com/improsocial/...
Might have to come out of (clinical psych) retirement for this one
After 5 years of data collection, our WARN-D machine learning competition to forecast depression onset is now LIVE! We hope many of you will participate—we have incredibly rich data.
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
WARN-D machine learning competition is live » Eiko Fried
If you share one single thing of our team in 2026—on social media or per email with your colleagues—please let it be this machine learning competition. It was half a decade of work to get here, especi...
eiko-fried.com
January 7, 2026 at 8:36 PM
Might have to come out of (clinical psych) retirement for this one
Reposted by Michael Mullarkey
After 5 years of data collection, our WARN-D machine learning competition to forecast depression onset is now LIVE! We hope many of you will participate—we have incredibly rich data.
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
WARN-D machine learning competition is live » Eiko Fried
If you share one single thing of our team in 2026—on social media or per email with your colleagues—please let it be this machine learning competition. It was half a decade of work to get here, especi...
eiko-fried.com
January 7, 2026 at 7:39 PM
After 5 years of data collection, our WARN-D machine learning competition to forecast depression onset is now LIVE! We hope many of you will participate—we have incredibly rich data.
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
If you share a single thing of my lab this year, please make it this competition.
eiko-fried.com/warn-d-machi...
If you like Huberman please do yourself a favor and go listen/watch Sad Boyz instead
More insight, less self-aggrandizement, and a far better parasocial hang
More insight, less self-aggrandizement, and a far better parasocial hang
January 7, 2026 at 3:48 AM
If you like Huberman please do yourself a favor and go listen/watch Sad Boyz instead
More insight, less self-aggrandizement, and a far better parasocial hang
More insight, less self-aggrandizement, and a far better parasocial hang
Reposted by Michael Mullarkey
serious question: is anyone hiring? open to connecting. 10+ years of full stack experience plus 1 year of unhinged ai/atproto hacking
January 6, 2026 at 9:52 PM
serious question: is anyone hiring? open to connecting. 10+ years of full stack experience plus 1 year of unhinged ai/atproto hacking
The best ways to not succumb to Poster's Madness
1. Never post
2. Have a good group chat
3. Literally anything else
1. Never post
2. Have a good group chat
3. Literally anything else
January 2, 2026 at 3:29 AM
The best ways to not succumb to Poster's Madness
1. Never post
2. Have a good group chat
3. Literally anything else
1. Never post
2. Have a good group chat
3. Literally anything else
In ~half of cases this fact brings me great relief. The other half have led to some of my greatest engineering frustrations
people joke about this but honestly half the time just doing the equivalent of a "unplug and plug back in" fixes things...
December 30, 2025 at 1:10 AM
In ~half of cases this fact brings me great relief. The other half have led to some of my greatest engineering frustrations
Evergreen
This is stupid. Going to the gym.
December 29, 2025 at 9:39 PM
Evergreen
My cat just set an unbreakable record for "fastest time to use the litter box after I cleaned it" by starting to use it while I was cleaning
Merry Christmas Eve everyone!
Merry Christmas Eve everyone!
December 24, 2025 at 11:08 PM
My cat just set an unbreakable record for "fastest time to use the litter box after I cleaned it" by starting to use it while I was cleaning
Merry Christmas Eve everyone!
Merry Christmas Eve everyone!