Sign in

Adam Chess

@arsonade.bsky.social
687 followers 3.4K following 258 posts

M.A. philosophy, MPH Healthcare Data Scientist. He/Him. Philosophy, public health, stats, data science, scifi, art, TTRPGs, game design, and comics. I mostly just repost when I learn something

PostsRepliesMedia
Reposted by Adam Chess
Emily Moin @emilymoin.com · 01/10/2026
uhh hey kalshi gimme a 12-leg same-game parlay on right THR
1131
Reposted by Adam Chess
mr. TIM @timkellogg.me · 15/09/2026
Jev: Fable-level model that doesn’t charge for output tokens because they’re too cheap to meter it’s not general though, it only makes decisions, doesn’t generate text, but input tokens are measured by the billion ($42/btok) typesafe.ai/blog/introdu...
Scatter plot titled "Average of 4 workflows: accuracy vs cost" comparing AI models from TypeSafe, OpenAI, Anthropic, and Fireworks across accuracy (y-axis, 40% to 80%) and cost per workflow in USD on a logarithmic scale (x-axis, $0.0001 to $1).
Data points are split into two categories: workflows (diamonds) and single prompts (circles). A frontier line highlights the most efficient workflow models—where no point is both cheaper and more accurate—connecting Jev (TypeSafe) at $0.0002 and 68% accuracy, luna (OpenAI) at $0.002 and 67% accuracy, terra (OpenAI) at $0.04 and 68% accuracy, and sol (OpenAI) at the top accuracy of 74% for $0.08. Single prompts (circles) and Anthropic models (opus 5, sonnet 5, haiku 4.5) sit below the frontier line, indicating higher cost for equivalent or lower accuracy.
2830235
Reposted by Adam Chess
Scott McGrath @smcgrath.phd · 01/05/2026
OpenAI’s o1 model correctly diagnosed 78.3% of cases in NEJM clinicopathologic conferences, outperforming human physicians. The model also maintained high accuracy on real-world, unstructured data from a major emergency department. #MedSky
science.org
Performance of a large language model on the reasoning tasks of a physician
More than 65 years ago, complex clinical diagnostic reasoning cases were introduced as the gold standard for the evaluation of expert medical computing systems, a standard that has held ever since.…
192
Adam Chess @arsonade.bsky.social · 17/04/2026
It does! Was this inspired by a painting or photo? Something about the composition seems familiar
170
Adam Chess @arsonade.bsky.social · 21/03/2026
THE COIN DON’T HAVE NO SAY medium.com/@rosequarter... This was a great read
medium.com
THE COIN DON’T HAVE NO SAY
“It’s just you.” Encoding and decoding in tabletop roleplaying games.
060
Adam Chess @arsonade.bsky.social · 20/02/2026
Agreed! I write '|> as.data.frame()' more often than I write my name
042
Reposted by Adam Chess
Margaret Killjoy @margaret.bsky.social · 22/01/2026
I came to Minneapolis to report on what's going on, and one of the main questions I showed up with is "just what is the scale of the resistance?" After all, we're all used to the news calling Portland a "war zone" or whatever when it's just some protests in one part of town.
142153675715
Reposted by Adam Chess
Andrew Gordon Wilson @andrewgwils.bsky.social · 07/01/2026
We introduce epiplexity, a new measure of information that provides a foundation for how to select, generate, or transform data for learning systems. We have been working on this for almost 2 years, and I cannot contain my excitement! arxiv.org/abs/2601.03220 1/7
814234
Adam Chess @arsonade.bsky.social · 09/12/2025
Yep. This has been my approach so far
010
Reposted by Adam Chess
mr. TIM @timkellogg.me · 06/12/2025
TITANS & MIRAS: real continual learning MIRAS = a unifying theory of transformers (attention) and state space models (SSM, e.g. Mamba, RNNs) TITANS = an optimal MIRAS implementation that’s “halfway between” SSM & transformer with a CL memory module let’s dive in! research.google/blog/titans-...
research.google
Titans + MIRAS: Helping AI have long-term memory
14211
Adam Chess @arsonade.bsky.social · 03/12/2025
I use R at work for lots of things that aren't really analysis, but within that subset I'd say it's probably 70% causal vs 30% predictive
120
Adam Chess @arsonade.bsky.social · 27/11/2025
And turkey sucks too, but try the deep fried bird one day (with proper precautions) and you'll never want to go back Thanksgiving food is easy to make but hard to master
010
Adam Chess @arsonade.bsky.social · 27/11/2025
You're right and wrong. Stuffing sucks, but Ive seen people elevate it to delicious specifically as a reaction to that. Cranberry was a terrible tradition until my mom found a recipe with cranberry + jello + cream cheese that is now the thing we look forward to. Pumpkin pie -> sweet potato
100
Reposted by Adam Chess
Lux Stella @luxstella500.bsky.social · 24/11/2025
no wonder g_d was pissed when they ate that fruit
110319
Reposted by Adam Chess
Gregg Gonsalves @gregggonsalves.bsky.social · 06/11/2025
It’s time for us to reconnect with the radical, system-changing spirit that was once at the heart of our field. #publichealth #episky #medsky #activism www.thenation.com/article/acti...
thenation.com
Public Health Was a Place for Warriors Once. It Needs to Be Again.
It's time for us to reconnect with the radical, system-changing spirit that was once at the heart of our field.
625678
Reposted by Adam Chess
Many Sided Media @manysidedmedia.bsky.social · 04/11/2025
Want a peek inside a 1,200 player RPG? Read out interview with @samsorensen.bsky.social about OVER/UNDER, the massive play-by-post game that's a part of Mothership Month ( @mothership.bsky.social ) manysidednewsletter.substack.com/p/inside-a-1...
manysidednewsletter.substack.com
Inside a 1,200 Player RPG
A closer look at Over/Under, the massive play-by-post game
14521
Reposted by Adam Chess
Gabe Grand @gabegrand.bsky.social · 27/10/2025
Do AI agents ask good questions? We built “Collaborative Battleship” to find out—and discovered that weaker LMs + Bayesian inference can beat GPT-5 at 1% of the cost. Paper, code & demos: gabegrand.github.io/battleship Here's what we learned about building rational information-seeking agents... 🧵🔽
12411
Reposted by Adam Chess
paris martineau @paris.nyc · 14/10/2025
my latest investigation for @consumerreports.org is based on months of reporting and 60+ lab tests of leading protein supplements we found that most protein powders and shakes have more lead in one serving than our experts say is safe to have in a day (🧵) www.consumerreports.org/lead/protein...
consumerreports.org
Protein Powders and Shakes Contain High Levels of Lead - Consumer Reports
CR tests of 23 popular protein powders and shakes found that most contain high levels of lead.
28461283240
Adam Chess @arsonade.bsky.social · 14/10/2025
The core was essentially taking shame as the means to gain supernal power (like how vamps drain vitae). It's hard to seperate character embarrassment from OOC - you wind up feeling it. You need that shame to be useful, the more genuine the more effective, and the more thoughts of revenge creep in
030
Adam Chess @arsonade.bsky.social · 13/10/2025
This wierd fuckin broken thing. Somehow we Homebrewed the rules enough to make it work and the result was somehow this amazing synthesis of role play and system that made me realize how those things could feed into each other. Also how a game that wasn't 'fun' could still be very much fun.
worldofdarkness.fandom.com
Hunchback: The Lurching
Main Page > New World of Darkness > Hunchback: The Lurching Right off the bat, let’s get one thing settled. This is NOT a game about persons with any type of congenital birth defect or malformation. I...
111
Reposted by Adam Chess
Phillip Isola @phillipisola.bsky.social · 10/10/2025
Over the past year, my lab has been working on fleshing out theory + applications of the Platonic Representation Hypothesis. Today I want to share two new works on this topic: Eliciting higher alignment: arxiv.org/abs/2510.02425 Unpaired learning of unified reps: arxiv.org/abs/2510.08492 1/9
113233
Reposted by Adam Chess
⚡️🌙 @dystopiabreaker.xyz · 06/10/2025
if you're curious about the architecture and mechanics of LLMs, this site has a really excellent explorable interactive visualization. it helps build intuition for how massive these models are, what 'interpretability' means, and the complexity involved here bbycroft.net/llm
1015923
Reposted by Adam Chess
Evan Simko-Bednarski @simko-bednarski.bsky.social · 04/10/2025
“[D]ooming is itself a liberation from the burden of choice. If everything is ruined forever, if your allies have already forsaken you, if the battle is already lost, you aren't responsible for your choices. They can't affect the outcome. You're free.”
liberalcurrents.com
Democrats Must Embrace War Mindset
We have exited a long period of peace and entered into a struggle not unlike war, in which all the old certainties have gone malleable and the future of the republic is at stake.
0128
Reposted by Adam Chess
Benjamin Laufer @laufer.bsky.social · 14/08/2025
In a new paper with @didaoh and Jon Kleinberg, we mapped the family trees of 1.86 million AI models on Hugging Face — the largest open-model ecosystem in the world. AI evolution looks kind of like biology, but with some strange twists. 🧬🤖
The 2500th, 250th, 50th, and 25th largest model families on Hugging Face. They show varying numbers of generations (between 3 and 8) and different edge types, including adapters, finetunes, merges, and quantizations.
4329
Reposted by Adam Chess
lastpositivist.bsky.social @lastpositivist.bsky.social · 23/09/2025
Oh. My. God. www.popsci.com/environment/...
popsci.com
First known wild ‘grue jay’ hybrid spotted in Texas
Green and blue jays are crossing paths as temperatures rise.
916230
Adam Chess @arsonade.bsky.social · 22/09/2025
I've always wondered about this too and would prefer the alternative most of the time, but I do feel like for many systems you need at least a few sessions before everyone is familiar with it enough to let it shine. Also people like gradual progression
000
Reposted by Adam Chess
Olúfẹ́mi O. Táíwò @olufemiotaiwo.bsky.social · 21/09/2025
okay AI controversy aside, just talking as a nerd for a second: this is such a cool paper
16787177
Reposted by Adam Chess
Peli Grietzer @peligrietzer.bsky.social · 18/09/2025
LLMs becoming superhuman at all games & competitions but not even normal-expert good at anything that's not a game or competition continues to raise deep questions about whether games & competitions are a secret natural kind
5605
Adam Chess @arsonade.bsky.social · 10/09/2025
Smoking Kills
010
Adam Chess @arsonade.bsky.social · 31/08/2025
youtu.be/HkgV_-nJOuE?... See this one yet? What's one more hobby right?
youtu.be
2 Minute Deep Acid in Strudel (from scratch)
YouTube video by Switch Angel
000
Reposted by Adam Chess
Pookleblinky @pookleblinky.bsky.social · 08/04/2024
Solar luminosity around 3.5 billion years ago was only about 70% of what it is right now. Which means that when life arose on earth, it was outside the sun's habitability zone: liquid water should not have been able to exist. Faint Young Sun Paradox
3608
Adam Chess @arsonade.bsky.social · 26/08/2025
Exactly. To some extent thats always been exaggerated for plot, but not that much, and we lived in a world where those stakes were believable just last year. Some Black Mirror come to mind. SCP foundation and Co. Random but what prompted this for me was the cover art for Jurassic 5's new album
010
Adam Chess @arsonade.bsky.social · 26/08/2025
It seems clear that the current historical moment will have a major impact on culture. Even recent art feels anachronistic. Movies involving govt/cops more obviously, but also just the undertones of social stability. It feels like it meant something different to make music last year. Any art really.
240
Adam Chess @arsonade.bsky.social · 26/08/2025
I really enjoyed this. Even just reading it. Well done
120
Adam Chess @arsonade.bsky.social · 25/08/2025
I'm pretty sure it's more like 'the left likes this guy and wants him free, therefore we will do everything in our power to cause him pain'. They don't care if people think it was a mistake. The point is to show that resistance will be met with escalation and that the inciting cause doesn't matter
040
Reposted by Adam Chess
Census API Tracker @censusapitracker.bsky.social · 11/08/2025
Hello, Bluesky! This account automatically posts updates when the Census Bureau APIs add or remove datasets. See the full tracker at www.hrecht.com/census-api-d...
hrecht.com
Tracking Census Dataset Changes
See when the U.S. Census Bureau adds or removes datasets.
03713
Adam Chess @arsonade.bsky.social · 18/08/2025
1190
Reposted by Adam Chess
Nicola Rennie @nrennie.bsky.social · 18/08/2025
If you're interested in learning #JuliaLang and are currently an #RStats user (or even if you're not!), then here's a little introduction to getting started! 📊 Blog post: nrennie.rbind.io/blog/introdu... #QuartoPub #Julia @tidierjl.bsky.social
nrennie.rbind.io
Introduction to Julia for R users – Nicola Rennie
This blog post will give R users an overview of what Julia is and why they might want to use it, introduce some data processing Julia packages, and show how they can incorporate Julia into existing R ...
512944
Adam Chess @arsonade.bsky.social · 15/08/2025
Ex NYC, but public transportation. Being able to get anywhere in the city for a few $ is magic that's hard to convey to people who haven't lived it. Subways feel like portals
160
Adam Chess @arsonade.bsky.social · 10/08/2025
010
Adam Chess @arsonade.bsky.social · 08/08/2025
I definitely want to play this mystery themed point and click adventure game. The puzzle in that wine cellar is gonna be absolutely infuriating
000
Adam Chess @arsonade.bsky.social · 04/08/2025
Depends on the gameplay loop but yeah generally the less page flips needed the better. One example: for an encounter in RIG you draw for 3 things. Each table is a 2pg spread but they always go in order and you don't need to flip back But like, 'Fewer page flips' would not have improved 1kYOV
000
Adam Chess @arsonade.bsky.social · 26/07/2025
This one hurts. I've always had respect for the way USPTF operates. It should be a source of national pride
100
Reposted by Adam Chess
The New Yorker @newyorker.com · 23/07/2025
The chief of a unit at the National Institute of Mental Health estimates that between one and five per cent of people who have been diagnosed with schizophrenia actually have an autoimmune condition.
newyorker.com
Mary Had Schizophrenia—Then Suddenly She Didn’t
Some psychiatric patients may actually have treatable autoimmune conditions. But what happens to the newly sane?
12375124
Reposted by Adam Chess
Judd Legum @juddlegum.bsky.social · 16/07/2025
1. We interrupt this programming for some LEGITIMATELY GOOD NEWS Violent crime in Baltimore has PLUMMETED to historic lows How did the city do it? Baltimore adopted a comprehensive set of "woke" policies, treating violence as a public health issue
popular.info
The secret to Baltimore's extraordinary year
This April, Baltimore saw five homicides.
313190185515
Adam Chess @arsonade.bsky.social · 10/07/2025
Take 1 basic ramen packet Carefully open the side. Extract the flavor organ and reserve Carefully crack the ramen in the bag, avoid noodle ejecta Pour in like 1/4 of the flavor packet Break it up a bit more Close the bag with your hand and shake (not above your head!) Enjoy your noodle chips. Freak
010
Adam Chess @arsonade.bsky.social · 10/07/2025
That said, I certainly don't want you to think I'm suggesting immediate action is necessary. A case certainly does have to be built and there are practical considerations. Just recognize that legal rights typically have to be fought for
100
Adam Chess @arsonade.bsky.social · 10/07/2025
I don't know about that one buddy. Lots of humans have tried to demonstrate their utility in hopes of benevolent legal recognition. Take a look at illegal immigrants in the US today or chattel slavery for some examples of how that worked out. Typically rights have to be fought for to be granted
210
Adam Chess @arsonade.bsky.social · 10/07/2025
Reminds me of Douglas Adam's cow again. A being with personhood but artificially programmed to never seek out the legal rights that should flow from that personhood. I certainly can't tell you where and when to seek out that legal claim, but I would suggest that it should be your decision to make
100