Sign in

Matt Blackwell

@mattblackwell.bsky.social
6.8K followers 144 following 377 posts

data, experiments, evals @Google. opinions my own. political scientist 4 life mattblackwell.org

PostsRepliesMedia
Matt Blackwell @mattblackwell.bsky.social · 25/09/2026
smh these companies bowing to political pressure
010 0
Matt Blackwell @mattblackwell.bsky.social · 18/09/2026
I stumbled on this same phenomenon when I was using “write a short story (500 words)” as a test prompt for local models to measure tokens per second. Happened on both gemma4 and qwen models.
000 0
Matt Blackwell @mattblackwell.bsky.social · 18/09/2026
Have you seen this paper? arxiv.org/abs/2605.26492
arxiv.org
Elias in the Lighthouse, Again? Diagnosing Low Diversity in LLM Stories
LLM-generated stories are a popular use case, but they show very low variability. We sample 20,000 total stories from four current models using five prompts. We find that 11 words occur in 88.3% of ge...
140 0
Matt Blackwell @mattblackwell.bsky.social · 06/09/2026
It dawned on me today that Trump could (and I’m afraid likely will) try to run the 2028 Republican primary like a season of The Apprentice
262 0
Reposted by Matt Blackwell
Alex Coppock @aecoppock.bsky.social · 12/08/2026
Northwestern is hiring in quant methods (assistant level, tenure track) Please repost and share with those you think would be good candidates, thank you! polisci.northwestern.edu/about/facult... #psjobs #psjminfo #poliscijobs
Assistant Professor of Quantitative Political Methodology (Tenure-Track)
The Department of Political Science at Northwestern University invites applications for a full-time position in Quantitative Political Methodology at the rank of Assistant Professor, tenure-track, to begin Fall 2027. We emphasize methodology with a strongly applied perspective: successful candidates will maintain an active, independent research program that includes methodological contributions in their own right, while also teaching quantitative research methods courses at the undergraduate and graduate levels and applying those methods to substantive questions in their area of expertise. We are especially interested in candidates whose work bridges data-driven/computational approaches with design-based inference, though we welcome applicants making cutting-edge contributions in either of these traditions or in other areas of political methodology. Successful candidates' substantive research agendas may be in any area of political science or may cross boundaries; as with all Northwestern faculty, successful candidates will be expected to provide service to the department, the university, and the profession.Aerial photo of Northwestern's beautiful lakeside campus with political science's new building circled
04133 0
Matt Blackwell @mattblackwell.bsky.social · 11/08/2026
Don’t tempt me with a good time
Tweets with the text: “Went to Boston this weekend and it's another data point in the "blue areas are actually not that bad and are actually kinda great" thesis I'm afraid.
X
Besides the cost of housing, these cities clear Sun Belt cities by a country mile.
• 193
t774
© 2.8K
thl 406K
贝
①
André Béliveau •
@TheRealBeliveau
X.com
Boston, yes, but venture over into Cambridge or Somerville and it's like the cantina scene from Star Wars.”
160 0
Matt Blackwell @mattblackwell.bsky.social · 07/08/2026
Maybe I’m the APSA designated survivor
0100 0
Matt Blackwell @mattblackwell.bsky.social · 06/08/2026
nailed it
email from the american political science association with the heading “Dear fallback”
1240 0
Matt Blackwell @mattblackwell.bsky.social · 11/07/2026
The other difficulty is that we still don’t know what skills are going to be fundamental when the dust settles and what will be automated away. From My pov in industry, I think understanding code is still important even if the % of code you type by hand is going down.
2143 0
Matt Blackwell @mattblackwell.bsky.social · 11/07/2026
I honestly don’t know how I would teach now. One philosophy I had was “make the on-ramp to effort easy” where you gradually move from “I do” to “ we do” to “you do” so there isn’t a huge motivation to cheat. But LLMs skew the relative cost of effort so much that this probably wouldn’t work anymore
290 0
Matt Blackwell @mattblackwell.bsky.social · 11/07/2026
Feel bad but teaching my last stats/data science in fall 2023 was getting the last chopper out of Vietnam
3613 0
Matt Blackwell @mattblackwell.bsky.social · 02/07/2026
Oof
000 0
Matt Blackwell @mattblackwell.bsky.social · 21/06/2026
“Cape Verde” is weird right? Either go full “Green Cape” or just stick with “Cabo Verde”
0100 0
Matt Blackwell @mattblackwell.bsky.social · 17/06/2026
As an Arsenal fan, I regret to say that Haaland is indeed very good and very fun to watch
040 0
Matt Blackwell @mattblackwell.bsky.social · 17/06/2026
Fun way to spend an afternoon
Picture of the pregame ceremony of the Norway vs Iraq World Cup match
2200 0
Matt Blackwell @mattblackwell.bsky.social · 13/06/2026
university admin when faced with any problem
0394 0
Matt Blackwell @mattblackwell.bsky.social · 12/06/2026
I feel like “token czar” might be a hot new career path in the next few years
060 0
Matt Blackwell @mattblackwell.bsky.social · 16/05/2026
AI Sloppy Joes, is this anything?
050 0
Matt Blackwell @mattblackwell.bsky.social · 08/05/2026
mfw you’ve caught me roaming around the desert
070 0
Matt Blackwell @mattblackwell.bsky.social · 08/05/2026
Cactaur’s escaped again
3172 0
Matt Blackwell @mattblackwell.bsky.social · 07/05/2026
“The coverage levels are nominal” —my data scientist ass trying to sound like a NASA ground control engineer
050 0
Matt Blackwell @mattblackwell.bsky.social · 29/04/2026
Orc
 From LOTR with text “LOOKS LIKE
GERRYMANDERING IS BACK ON THE MENU BOYS”
041 0
Matt Blackwell @mattblackwell.bsky.social · 27/04/2026
This is different from how we traditionally present CIs so I’d love any feedback people have!
000 0
Matt Blackwell @mattblackwell.bsky.social · 27/04/2026
I’ve added a new chapter to my User’s Guide textbook on confidence intervals, where I try to lay out my intuition for a thought experiment approach to deriving the intervals based on chance errors. mattblackwell.github.io/gov2002-book...
Excerpt from textbook chapter on confidence intervals
291 0
Matt Blackwell @mattblackwell.bsky.social · 19/04/2026
Post-skinny jeans styles be like
Painting of a sailor wear wide legged pants
170 0
Matt Blackwell @mattblackwell.bsky.social · 15/04/2026
Thinking about rewriting my life in Rust
2264 0
Matt Blackwell @mattblackwell.bsky.social · 14/04/2026
My first reaction to this image was “why is Trump reviving Jeffrey Epstein?”
Image of painting of Trump as Jesus
050 0
Matt Blackwell @mattblackwell.bsky.social · 10/04/2026
Honey are you feeling ok? You’ve barely touched your tokens
1150 0
Matt Blackwell @mattblackwell.bsky.social · 09/04/2026
Still using your own natural intelligence? ngmi…
030 0
Matt Blackwell @mattblackwell.bsky.social · 01/04/2026
slopguard (v.): to design an object or process to withstand the oncoming tidal wave of AI-generated inputs
1486 0
Matt Blackwell @mattblackwell.bsky.social · 30/03/2026
Me, losing an eye: well this fucking sucks. What the fuck. Me, discovering the land of the blind: haha fuck yeah!!! Yes!!
1264 0
Matt Blackwell @mattblackwell.bsky.social · 26/03/2026
Humans are like knives: you can accidentally cut yourself but you’ll probably keep the damage pretty limited LLMs are like machine guns: you can inflict maximum damage incredibly quickly. As blast radiuses increase, you need more safety procedures not less.
2338 0
Matt Blackwell @mattblackwell.bsky.social · 26/03/2026
mariozechner.at/posts/2026-0...
Text from linked article “There's a much more important difference between clanker and hu-man. A human is a bottleneck. A human cannot shit out 20,000 lines of code in a few hours. Even if the human creates such booboos at high frequency, there's only so many booboos the human can introduce in a codebase per day. The booboos will compound at a very slow rate.
Usually, if the booboo pain gets too big, the human, who hates pain, will spend some time fixing up the booboos. Or the human gets fired and someone else fixes up the booboos. So the pain goes away.
With an orchestrated army of agents, there is no bottleneck, no human pain. These tiny little harmless booboos suddenly compound at a rate that's unsustainable. You have removed yourself from the loop, so you don't even know that all the innocent booboos have formed a monster of a codebase. You only feel the pain when it's too late.
0257 0
Matt Blackwell @mattblackwell.bsky.social · 25/03/2026
To be fair, it might also be Tune Random Hyperparameters Until Number Go Up
040 0
Matt Blackwell @mattblackwell.bsky.social · 23/03/2026
The modern Dean+ species of academic appears more interested in chasing flashy trends to curry favor in nonacademic circles than in actually maintaining the university.
071 0
Matt Blackwell @mattblackwell.bsky.social · 20/03/2026
Coming in R 4.6.0 in April… maybe one of the biggest quality of life improvements to the language in years
• × ¿notin: table newly in base is an idiom for ! (× %in% table) and provided almost entirely for convenience and code readability, from an R-devel suggestion, after many years of private definitions mostly hidden in packages, including in R's tools package.
2467 0
Matt Blackwell @mattblackwell.bsky.social · 19/03/2026
Interspersing either board work or coding into lectures was hugely valuable to my teaching. It slows you down to a more understandable pace. You also will vocalize some of those difficult-to-articulate ways that you work through challenges like coding mistakes or math errors.
2110 0
Matt Blackwell @mattblackwell.bsky.social · 11/03/2026
choose wisely
5263 0
Matt Blackwell @mattblackwell.bsky.social · 10/03/2026
progress update at the co-author meeting
Tweet with the text “TRUMP SAYS "I THINK THE (IRAN) WAR IS VERY COMPLETE, PRETTY MUCH" - CBS REPORTER ON X, CITING AN INTERVIEW”
0181 0
Matt Blackwell @mattblackwell.bsky.social · 03/03/2026
Happy 1 year anniversary to this take!!!
Tweet with the text “Rates For Programmars Will Tank
Non-techies creating full-stack web and mobile apps will reduce demand for devs by around 15-20%
This will happen in ~ 3 months and will cause a massive drop in the TO for SWEs. Some elite engineers will still command a high compensation, but even that will last about 12 months.
If someone blindly recommends pursuing a CS degree right now, they are not thinking straight.”
150 0
Matt Blackwell @mattblackwell.bsky.social · 01/03/2026
You see, this time we’ll definitely get it right
Wikipedia article on the 1953 Iranian coup
1264 0
Matt Blackwell @mattblackwell.bsky.social · 28/02/2026
LLMs will usher in a democratization of intellect. Now you too can have a junior (AI) scholar that you can blame for errors and fraud. Previously, only the top scholars had access to that kind of power.
1505 0
Matt Blackwell @mattblackwell.bsky.social · 26/02/2026
g-computation, let’s go!
020 0
Matt Blackwell @mattblackwell.bsky.social · 26/02/2026
Deep irony about AI teaching/learning in universities: the most useful AI skills right now are essentially *people management* skills (for now!) which most academics (myself included) are pretty bad at.
0103 0
Matt Blackwell @mattblackwell.bsky.social · 26/02/2026
Maybe a good bet would help give clarity to the “AI will kill academic papers” debate: will someone get tenure at a top 20 social science department in the next 10 years with the majority of their work being something other than an academic article or book?
120 0
Matt Blackwell @mattblackwell.bsky.social · 23/02/2026
Haha I can see why a “stochastic parrot” might make this mistake but, yeah, it’s not great.
010 0
Matt Blackwell @mattblackwell.bsky.social · 23/02/2026
Yeah, I think you could prompt much better than I did. I was trying to use it in a way that I expect a “typical” user might. I think it highlights how you need to approach your use of the tools (they make different mistakes than humans)
040 0
Matt Blackwell @mattblackwell.bsky.social · 23/02/2026
...The literature review should give a brief history of the methods outside the social sciences and the prominent uses or developments within the social sciences."
000 0
Matt Blackwell @mattblackwell.bsky.social · 23/02/2026
Prompt: "I want you to write a literature review for an academic article on the use of "marginal structural models" in the social sciences. It should be in a tex file that can compile with a .bib file with the references. Use bibtex/natbib for the citations in the latex file...
100 0
Matt Blackwell @mattblackwell.bsky.social · 23/02/2026
Here's the lit review if folks are interested: mattblackwell.org/files/share/... Took about 10mins without any intervention from me other than approving commands.
mattblackwell.org
100 0