Sebastian Karcher
LightNews LightNews
banner
adam42smith.bsky.social
Sebastian Karcher
@adam42smith.bsky.social
1.2K followers 760 following 1.8K posts

Director, Qualitative Data Repository (personal account).
Data, Zotero, Social Science Methods
https://sebastiankarcher.com

Videos
Political science 25%
Computer science 23%
Posts Replies Media Videos

Reposted by Ingo Rohlfing, Sebastian Karcher

kunkakom.bsky.social
Alexander Wuttke @kunkakom.bsky.social · 9h
The sale pitch for this AI Scientist "Kosmos" as presented in this Podcast just seems like a big HARKing exercise. Yes, just look at the data long enough and you'll find "something".

Has anybody made any experiences with AI scientists models? Recommendations?
pca.st/episode/58aa...
Where Is All the A.I.-Driven Scientific Progress?
pca.st

Reposted by Sebastian Karcher, Joanna Tai, Jesús García

xkcd.com
Randall Munroe @xkcd.com · 1d
Funny Numbers

xkcd.com/3184/
Comic. [Banner: Mathematical Society: 2025 Meeting.] PERSON 1 with bun: Any other new developments from the year to cover before we wrap? PERSON 2: Oh, the teens picked a new funny number. PERSON 3 with short hair: Aww, I’m glad to hear they’re still doing that. PERSON 4 with ponytail: I’ll add it to the list. [List: 23 (skidoo!; 42; 69; 420; 1,337; 58,008; [circled]: 67]

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 1d
Vegans hold all sorts of different views, but Vegans in the Singer/Animal Rights tradition think pretty much exactly that - they'd even use the term 'specieist'. Contra stereotype, the vast majority of them also understand that yelling at people who disagree is both obnoxious & counterproductive

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 2d
Absolutely wild. It was crazy even before it got to the Boccelis...

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 2d
If you look at the submitted/accepted dates, I think we can conclusively say this wasn't peer-reviewed, so the label is wrong.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 2d
Skocpol is a brilliant scholar & sharp political analyst who also suffers from 'kids these days' syndrome, including 'kids these days & their unions' npeu.org/news/2020/7/...

(Also, the though I can't find it anymore, kids these days & their sexual harassment concerns)

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 2d
I think what it's supposed to be is lower Eastside but maybe some deep subconscious kept him from blaming NY Jews.
Alternatively, CBS producers are really into theater & have second apartments in London...

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 3d
FWIW you can just cancel your Zotero subscription and keep everything on your computer

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
I was just remarking on the plethora of hip coffee shops and wine bars in Pine Falls (which also has a surprisingly large Jewish congregation bc we're watching Oy to the World)

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
Fascinating. I'm surprised this particular problem seems so hard to solve.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
I don't think your initial 'improve your writing' framing works (as opposed to fleshing out rough notes as in the paper figure).
a) If you give AI fully crafted writing to improve, IME feedback rarely introduces new ideas
b) if it did, no reason you wouldn't notice it needed a citation

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
The journal doesn't seem to have a 'letter' or similar category, but presumably really wanted to publish it. Most generously, because they thought it was important, but hard to think it wasn't also because of the names involved & bc of what it'll do for their IF...

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
that should have been "less than two months". ugh.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 4d
This was published >2 months after initial submission as a "Research Article" in journal that claims articles are peer reviewed. I'm all for rapid turnaround, but this is hard to believe.
If they want to publish rapidly on important issues w/o peer review -- great, but should be labelled as such

Reposted by Sebastian Karcher

adamcsharp.bsky.social
Adam Sharp @adamcsharp.bsky.social · 4d
In one of the Danish TV screenings of Die Hard, “yippee-ki-yay, motherfucker” was changed to “yippee-ki-yay, bror lort,” which means “yippee-ki-yay, shit brother.” In a German screening, the “motherfucker” was changed to Schweinebacke (“pig cheek”). And in Finnish... pulunnussija (“pigeon fucker”).

Reposted by Anna O. Law, Richard Waite, Sebastian Karcher , and 1 more

ruthkapelus.bsky.social
Ruth Kapelus @ruthkapelus.bsky.social · 4d
Let’s air those grievances. I want to hear your pettiest gripes on this the holiest of days.

Reposted by Sebastian Karcher, Nikki Fairchild

ginnybraun.bsky.social
Prof. Virginia Braun (she/her they/them) @ginnybraun.bsky.social · 4d
Now published…

We Reject the Use of Generative Artificial Intelligence for Reflexive Qualitative Research - Tanisha Jowsey, Virginia Braun, Victoria Clarke, Deborah Lupton, Michelle Fine, 2025

journals.sagepub.com/doi/10.1177/...
We Reject the Use of Generative Artificial Intelligence for Reflexive Qualitative Research - Tanisha Jowsey, Virginia Braun, Victoria Clarke, Deborah Lupton, Michelle Fine, 2025
Four hundred and nineteen experienced qualitative researchers from 32 countries invite readers of Qualitative Inquiry to consider their position on use of gener...
journals.sagepub.com

Reposted by Anna O. Law, Rosemary A. Joyce, Rebecca Tushnet , and 10 more

appuhn.bsky.social
Karl Appuhn @appuhn.bsky.social · 5d
I think we need a mega thread of everyone's craziest archive stories.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
2/2 It does mean that you may make an invalid ILL request, which libraries already report is happening a lot :(.
I see how that's a hassle for libs & users, but this was never a good way to verify citations; that seems like something teachable (&obv a citr isn't actually verified until you see PDF)

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
What's happening is that the library discovery layer is displaying the output of an OpenURL request. It does the same, e.g., for Zotero's "Locate" function. That's extremely useful bc you can then a) either access the work via the library or b) make an ILL request 1/2

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
To be clear, though: you only get to that pseudo catalog page if you start on GScholar (and the functionality isn't GScholar specific). If you start in the library catalog, you don't see the cited 'work' existing ever

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
I see people are applying to be main character of Bsky like in the good ol days of yore.
I preferred this place without.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
In the specific case of football, though, that confidence sounds very painful...

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 6d
I had somehow never seen that graph -- that's incredibly effective
michelnivard.bsky.social Michel Nivard @michelnivard.bsky.social · 7d
"Across the highest adult performance, peak performance is negatively correlated with early performance" Is uite literally what you'd see if coliderbias:

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 7d
Just for the ratio/selection bias: I'm with you on this. The companies suck, some of the societal consequences large & small are v worrying, but the tech is amazing

Reposted by Sebastian Karcher

michelnivard.bsky.social
Michel Nivard @michelnivard.bsky.social · 7d
"Across the highest adult performance, peak performance is negatively correlated with early performance" Is uite literally what you'd see if coliderbias:

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 7d
A-Index. 42 is quite the humble-brag, I have to say.

Reposted by Anna O. Law, Alan Richardson, Sebastian Karcher

dieworkwear.bsky.social
derek guy @dieworkwear.bsky.social · 8d
the epstein files as rick owens fall/winter 2024
Two photos. One shows a page from the Epstein files. Almost the entire page has been blacked out except for a single line at the top. The second photo shows an all-black Rick Owens outfit with the model's white face being the only visible part. Two photos. One shows a page from the Epstein files. Almost the entire page has been blacked out except for two lines at the top and five lines at the bottom. The second photo shows an all-black Rick Owens outfit with the model's white face and legs being the only visible parts. Two photos. One shows a page from the Epstein files. The entire page has been blacked out. The second photo shows an all-black Rick Owens outfit with the black model's face being the only visible part. Two photos. One shows a page from the Epstein files. Almost the entire page has been blacked out except for a block of text at the top. The second photo shows an all-black Rick Owens outfit with the model's white face being the only visible part.

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 8d
The NYT very oddly vacillates between calling lies lies and weaseling around them. They absolutely know how to call out lies, e.g.:
(source: www.nytimes.com/2025/12/19/n... )
A lengthy passage from an article about Elise Stefanik that has the fragments "amplifying his lies about the 2020 election" and "who she falsely claimed supported" highlighted

adam42smith.bsky.social
Sebastian Karcher @adam42smith.bsky.social · 8d
Looking for #polisky data -- I'm looking for a recent-ish paper with public survey data that has an English free text response to at least one Q that, ideally, is categorized and analyzed in the paper. (Background: I'd like to experiment w some classifiers, likely for teaching)
Any pointers/ideas?