Sign in

Stephen Wild

@stephenjwild.bsky.social
4K followers 1.1K following 11K posts

I try to put straight lines through things but usually fail. Try to be Bayesian when I can. Views my own. RT/like != endorsement.

PostsRepliesMedia
Stephen Wild @stephenjwild.bsky.social · 2h
NOW I AM HAUNTED 😭
130
Stephen Wild @stephenjwild.bsky.social · 7h
Just wear a shirt jacket and claim to be a working man.
010
Stephen Wild @stephenjwild.bsky.social · 7h
www.upress.umn.edu/978151791932...
upress.umn.edu
Language Machines
How generative AI systems capture a core function of language Looking at the emergence of generative AI, Language Machines presents a new theory of meaning i...
020
Stephen Wild @stephenjwild.bsky.social · 9h
I would be happier if it was something like "requires minimal parametric assumptions in comparison to [method]" or some such. Alas. bsky.app/profile/eric...
130
Stephen Wild @stephenjwild.bsky.social · 9h
As a Steve I concur. Steve and intelligence do not go together.
030
Stephen Wild @stephenjwild.bsky.social · 9h
I clicked through but haven't read them yet. I'm mostly annoyed by the continual use of the term "guaranteed".
130
Stephen Wild @stephenjwild.bsky.social · 9h
Conformal prediction is one of those ideas that seems much more complicated than it is, but once you understand it you immediately see why language liked "guaranteed coverage" should be avoided
130
Stephen Wild @stephenjwild.bsky.social · 9h
Anyway it's a good post
220
Stephen Wild @stephenjwild.bsky.social · 9h
Not the point of this paragraph, but "conformal prediction" is one of those things that makes me uneasy because I dislike its claims of guaranteed coverage* *not be a guarantee
And that’s more or less all of the tricks we know. How many classes must we spend on it? How many papers must we write about it? [footnote: A Google Scholar search of “conformal prediction” says the answer is thousands.] In this class, we’ll only use one.
350
Stephen Wild @stephenjwild.bsky.social · 9h
A cheap Garmin is cheaper than an Apple watch. But also wait for the sale.
130
Stephen Wild @stephenjwild.bsky.social · 11h
"So I need a bigger sample?" "No, no, I mean you need..." "A bigger sample." "No, no. That's not what I mean at all." "A bigger sample is all I need..."
120
Stephen Wild @stephenjwild.bsky.social · 12h
Exquisite and infuriating bait.
1160
Stephen Wild @stephenjwild.bsky.social · 12h
Had I waited to respond I would have seen that we agree bsky.app/profile/grim...
010
Stephen Wild @stephenjwild.bsky.social · 12h
I think simulating responses to see if your survey/questions/etc work is fine (and a perfectly good use case). My issue is with treating the responses as if they weren't simulated.
140
Stephen Wild @stephenjwild.bsky.social · 12h
Just don't.
130
Stephen Wild @stephenjwild.bsky.social · 12h
A meme with two panels. In panel 1 the woman looks disgusted. In panel 2 intrigued.

Text, panel 1: fake data.

Text, panel 2: silicon sample
2200
Stephen Wild @stephenjwild.bsky.social · 13h
This, but instead it was Goodnight, Goodnight, Construction Site bsky.app/profile/ldob...
130
Stephen Wild @stephenjwild.bsky.social · 13h
Absolutely ridiculous. [String of profanity here]. Yes, I am taking a reddit post at face value because I have seen this dynamic
100
Stephen Wild @stephenjwild.bsky.social · 13h
static.klipy.com
Homer Simpson's Loudest Scream
Alt: Homer Simpson's Loudest Scream. He says the F-word, and yells it so loudly everyone in the town hears.
180
Stephen Wild @stephenjwild.bsky.social · 14h
Though now I am going to look up Blood Oil
110
Stephen Wild @stephenjwild.bsky.social · 14h
Just realized I misread who it was. My bad.
110
Stephen Wild @stephenjwild.bsky.social · 14h
I'm one chapter into his book right now. So far it's good (albeit dense)
110
Stephen Wild @stephenjwild.bsky.social · 15h
Recommend something for Matt, nerds!
012
Stephen Wild @stephenjwild.bsky.social · 16h
Hoff can be helpful, bit still a bit of a slogan. sites.math.rutgers.edu/~zeilberg/EM...
sites.math.rutgers.edu
010
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Damn, that sucks
010
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Seriously, though, the evidence base for a lot of meds (especially older ones) is pretty slim
030
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Jeff Goldblum voice: Based on the studies I usually see on here the answer is--uh--not much
120
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Honestly, it was probably the bag of candy I ate that helped get rid of it
110
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
You misspelled "deeply flawed" 😉
040
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Yes sickos journals.sagepub.com/doi/abs/10.1...
journals.sagepub.com
Probing Birth-Order Effects on Narrow Traits Using Specification-Curve Analysis - Julia M. Rohrer, Boris Egloff, Stefan C. Schmukle, 2017
The idea that birth-order position has a lasting impact on personality has been discussed for the past 100 years. Recent large-scale studies have indicated that...
180
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
The "What's Hot?" Feed was a mistake
020
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
How I arrived on Bluesky (IYKYK)
static.klipy.com
Flanders and Homer Skiing in Flandersski Suit
Alt: Flanders and Homer Skiing in Flanders ski Suit. "Stupid sexy Flanders"
170
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Help out tiddlydump and Mercer!
010
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
It's happening again
3323
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
It's air conditioning for where Andrew lives
020
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Because I don't know anything about stratification sociology, behavioral psychology, and institution political science
020
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Sociology, psychology, poli sci...: all known for their use of expensive equipment, proofs, and lack of synthesis
120
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
do other social science disciplines exist send tweet
We associate the higher concentration of prizewinning work in economics with the field’s stronger sorting by institutional prestige, its lower reliance on specialized equipment and instruments, and its assessment of findings based on a synthesis of evidence rather than on decisive experiments or proofs.
161
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
(Not quite what happened in the case of Bacon's Rebellion, but a similar dynamic)
020
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
A fairly common theme is: government negotiates treaty with indigenous tribes (we'll pretend it's above board--they often weren't), settlers get angry and ignore treaty, government lacks manpower and authority to enforce it, indigenous tribes fight back, settlers attack, repeat process.
120
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
But was it really a revolution? bsky.app/profile/adam...
010
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
There are lots of reasons for FN to be skeptical of the Crown and governments in general, given the events of the subsequent 260+ years.
060
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Semi-fun fact but a primary impetus for the American Revolution was the British Crown protecting aboriginal lands from encroaching states.
3202
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Excuse me, but this movie was clearly aimed at me, a 20ish year old man when the movie came out. I loved it. Also Liam is correct.
030
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
I'm in Canada. We use Fahrenheit, like nature intended
130
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
He's not in Germany so it's fine
120
Reposted by Stephen Wild
Cameron Patrick @cameronpat.bsky.social · 07/10/2026
of all the methods for dealing with missing data, "keep asking until someone finds it, or finds out why it wasn't recorded" rarely appears in textbooks but can be SUPER EFFECTIVE in practice
2274
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Shhh. Don't tell them what cucumbers are
000
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
I am still focused on that list of names. 4 of them are not who I would aspire to be (I have no idea who the other one is)
120
Stephen Wild @stephenjwild.bsky.social · 07/10/2026
Okay at least it's salvagable.
010