iiinova — where your digital self keeps evolving

Vilgot Huhn @vilgothuhn.bsky.social · 10/10/2026

Perhaps for Americans. I've seen on the TV about your "PBs and J" sandwich. Here in the EU, regulation recommends a maximum 2 jars.

110

Vilgot Huhn @vilgothuhn.bsky.social · 10/10/2026

You can even take it one level further: I want to want to want to quit smoking (but I don't so I don't so I don't)

010

Vilgot Huhn @vilgothuhn.bsky.social · 10/10/2026

it never registers that we already have peanut butter so we always end up with 3-5 jars

100

Vilgot Huhn @vilgothuhn.bsky.social · 10/10/2026

why does stuff like this excite me?

130

Reposted by Vilgot Huhn

Julia M. Rohrer @dingdingpeng.the100.ci · 10/10/2026

Wake up babe, new polyhedra facts just dropped! All noble polyhedra have been enumerated…by a 17-year-old. m.youtube.com/watch?v=9533...

m.youtube.com

Big News in Polyhedra - Numberphile

YouTube video by Numberphile

0274

Reposted by Vilgot Huhn

Grace @gracekind.net · 09/10/2026

Everything is computer, except computer, which is power

720126

Vilgot Huhn @vilgothuhn.bsky.social · 08/10/2026

Definitely! I don't mean to imply we should never be concerned with unnecessary duplicates of already existing concepts. Indeed, I think that's a quite big problem and it makes psychology a messy field. I just think there's another extreme which is also bad.

000

Reposted by Vilgot Huhn

Colin @colin-fraser.net · 08/10/2026

Calculators are impressive

317812

Vilgot Huhn @vilgothuhn.bsky.social · 08/10/2026

Some people use it to be rigorous critical thinkers, or something. But it gives people a very impoverished view of humanity. And I don't just mean that in some romantic sense. If your bar for including something in your model of psychology is too high, at some point you're just close-minded.

200

Vilgot Huhn @vilgothuhn.bsky.social · 08/10/2026

There's this "deflationary" version of psychology where everything reduces to a few things. Every cognitive/executive trait is just intelligence. All human behavior just comes down to incentives. All psychological disorders are just neuroticism. I don't like it.

110

Vilgot Huhn @vilgothuhn.bsky.social · 07/10/2026

History happens once. First as tragedy.

000

Vilgot Huhn @vilgothuhn.bsky.social · 07/10/2026

I'm just going on titles in the list of highly influential papers but "Psychological language on Twitter predicts county-level heart disease mortality" sounds Wild! Looked into it and alas, there has already been debate/commentary about that. By @steamtraen.eu if I'm not mistaken

sciencegate.app

140

Vilgot Huhn @vilgothuhn.bsky.social · 07/10/2026

My take is that methodology thrives well on social media because: 1. common ground for all researchers. 2. often has exciting conflict (replication failures, bayes vs frequentism). 3. let's you feel smarter than others.

020

Vilgot Huhn @vilgothuhn.bsky.social · 07/10/2026

I've been idly fantasizing about organizing debates on substantive issues in psychology. Discussed this within my research group and they seemed interested in it, but so far the only topic that has been suggested is methodology related. The lure of method is just too strong!

120

Reposted by Vilgot Huhn

SMBC Comics @smbccomics.bsky.social · 01/10/2026

That janitor just goes from building to building helping people. COMIC ◆ www.smbc-comics.com/comic/equati... PATREON ◆ www.patreon.com/ZachWeinersm... STORE ◆ smbc-store.myshopify.com

10-panel SMBC comic update. All the panels are dialogue-less. Two researchers are writing equations on a chalkboard. A janitor watches them from afar. The researchers leave as the janitor is still mopping the floor. The janitor walks up to the chalkboard and writes on it with purpose, admires what he wrote, and leave late in the evening. The two researchers return at dawn, and are shocked by what is written on the blackboard. Over their half-erased equations, the text "ChatGPT, do a breakthru, no mistakes, thankx."

1044682

Vilgot Huhn @vilgothuhn.bsky.social · 07/10/2026

I live with a constant shame about not really knowing math. In one way these latest developments could be taken as proof I actually chose the right path in life, but I don’t know… All this fills me with a melancholy more than anything.

000

Reposted by Vilgot Huhn

hailey @hailey.at · 06/10/2026

lmao

mistral leading on on a legal agent benchmark

822615

Reposted by Vilgot Huhn

Ryan Moulton @moultano.bsky.social · 06/10/2026

The Shepard tone of climatology is constantly redoing your color scale to accommodate the new data.

3418

Reposted by Vilgot Huhn

joe ⧉ 🍷🌊 @joetforhire.com · 06/10/2026

prowling up and down the feed with my new reaction image in hand

21199

Reposted by Vilgot Huhn

Nick Rempel @nrempel.com · 06/10/2026

wild that this is exactly what using a computer is like now

1037380

Vilgot Huhn @vilgothuhn.bsky.social · 05/10/2026

@pgmj.bsky.social @rpsychologist.com @matti.vuorre.com might be interested since you guys recently engaged with this paper

120

Vilgot Huhn @vilgothuhn.bsky.social · 05/10/2026

Update on the Hayes et al. "Call for an idionomic revolution" AI citation drama. They've now published a correction. "The published reference does not correspond to a published work." Three times. And some other stuff. www.sciencedirect.com/science/arti...

sciencedirect.com

130

Vilgot Huhn @vilgothuhn.bsky.social · 03/10/2026

My career plan B is to just take cognitive therapy, replace the words schema/beliefs with the words model/priors, and sell it as a revolutionary new thing. Sprinkle in some neuroscience. I’m talking workshops, certificates, maybe even an institute. ”Bayesian Brain Therapy”

010

Vilgot Huhn @vilgothuhn.bsky.social · 01/10/2026

Well they don’t have a pineal gland so how would their res exstensa interact with any res cognitans?

010

Vilgot Huhn @vilgothuhn.bsky.social · 30/09/2026

Also: Weird to see that some people appear to become pro AI simply because of 3. "...so preoccupied with whether or not they could, they didn’t stop to think if they should" and all that. How will this change society? Whatever, let's hope for the better, but most of all I just wanna build!

000

Vilgot Huhn @vilgothuhn.bsky.social · 30/09/2026

I guess my ambivalent feeling can most clearly be described as "anti-pro-AI". More than anything, I do not relate to people who feel no sense of dread, or at least resistance to change. But that doesn't mean I think no one should use the useful little guy in the computer.

120

Vilgot Huhn @vilgothuhn.bsky.social · 30/09/2026

Your Fallible Human Eyes Are Not Better At Detecting Correlation Than OLS

001

Vilgot Huhn @vilgothuhn.bsky.social · 30/09/2026

I’m more focused on the object-level issue here: Do geese see God?

000

Vilgot Huhn @vilgothuhn.bsky.social · 29/09/2026

The thing about RM-ANOVA is that it seems somehow *more* difficult to use than mixed models while also being all around worse? Am I missing something here?

130

Vilgot Huhn @vilgothuhn.bsky.social · 28/09/2026

man I hope the real treasure isnt the friends we made along the way because AI is getting very effective at skipping past the journey

000

Reposted by Vilgot Huhn

Magnus Johansson @pgmj.bsky.social · 28/09/2026

Our commentary "A Revolution in Search of an Estimand: On the Idionomic Proposal of Hayes et al. (2026)" is now available as a preprint: osf.io/preprints/ps... With @rpsychologist.com, @matti.vuorre.com, and Nils Hentati Isacsson.

osf.io

OSF

03218

Vilgot Huhn @vilgothuhn.bsky.social · 28/09/2026

I’m usually pretty sympathetic to the idea that we shouldn’t anthropomorphize LLMs but come on! ”Tried”? ”misbehaved”? I’ll say ”my phone tried to connect to the wifi”, that doesn’t mean I think it’s a little guy.

000

Vilgot Huhn @vilgothuhn.bsky.social · 27/09/2026

Same thing with Fable, those few days I had access. The coding capabilities I understand is a leap, but as a conversation partner I though it was, if anything, worse than Opus

020

Vilgot Huhn @vilgothuhn.bsky.social · 27/09/2026

To be clear, I did not feel like this when I went from free Sonnet to paid Opus 4, even outside of claude code. That was a jump. That was the difference between hallucinatory circling conversations whenever I chatted about something complicated, to genuinely insightful stuff.

140

Vilgot Huhn @vilgothuhn.bsky.social · 27/09/2026

Obviously I'm not a very expert LLM user (at least these last few months) because just based on chatting/small tasks Opus 5.5 is like Opus 5. I can't tell the difference. I've seen the demonstrations. Super intimidating. But day to day for me it's the same chatbot as yesterday.

140

Vilgot Huhn @vilgothuhn.bsky.social · 24/09/2026

Amazingly pedagogical ("pedagogiskt", säger man så på engelska?) Thank you for this one! Although I wouldn't push the reasoning too far. In some contexts something like "within-strata association" is a fine descriptive research question, right?

100

Reposted by Vilgot Huhn

Kristoffer Magnusson @rpsychologist.com · 25/06/2026

New interactive blog! "Why Adjusted Regression Coefficients Are Less Descriptive Than They Look" rpsychologist.com/descriptive-...

1520769

Vilgot Huhn @vilgothuhn.bsky.social · 24/09/2026

Even so, if it were up to me I would say no more. Let's do Opus 4.5/4.6, whichever one was least obnoxious, and leave it there. That's enough for me, thanks. Apply that tech where ever you want and lets not break the world with these things.

020

Vilgot Huhn @vilgothuhn.bsky.social · 24/09/2026

How do networks of neurons give rise to meaningful information processing/behaviour? That's just inherently fascinating to me. I'm sure brains don't work quite like transformers but the fact that LLMs do something quite "deep" is surely a related inquiry.

110

Vilgot Huhn @vilgothuhn.bsky.social · 24/09/2026

Access to the third state gives my emotions a little break. I used to be a huge neuro nerd as an undergrad. I read brain books way above my level. Eventually I got sad about how that didn't really lead anywhere for me. But even so, the brain is the most interesting thing there is. How the hell?

110

Vilgot Huhn @vilgothuhn.bsky.social · 24/09/2026

With AI I'm constantly cycling between: 1. This tool is amazing! Claude is such a useful little guy. 2. This is monstrous! The death of art. Death of purpose. Get me of this wild ride. I hate this. 3. The fact that LLMs work this well is just deeply fascinating and bizarre and I need to know more.

131

Vilgot Huhn @vilgothuhn.bsky.social · 23/09/2026

Wait, Adam Sandler’s Click (2006) was prophetic? I hate this timeline.

000

Reposted by Vilgot Huhn

paolo @paolopalma.bsky.social · 20/09/2026

Working on a new taxonomy for research methods

Punt square with rows depicting data collection (fun, not fun) and columns depicting data analysis (fun, not fun)

Qual interviews are listed as fun to collect and not fun to analyze while surveys are listed as fun to analyze but not fun to collect

55514

Vilgot Huhn @vilgothuhn.bsky.social · 20/09/2026

I had the freakout but not sure I integrated it. Never resolved it. Left it open on a todo list. I think that experience is very common: "I'm my brain and brains are physical? Scary! Does that mean one can build some conscious wooden Rube Goldberg machine? Intuitively no. Let's leave it there."

020