Antoine Vernet
@antoinevernet.com
Associate Professor in Management at UCL.
Networks and Organizational Design.
Mostly found reading.
Very occasionally on Youtube: www.youtube.com/@antoinevernet
Networks and Organizational Design.
Mostly found reading.
Very occasionally on Youtube: www.youtube.com/@antoinevernet
Reposted by Antoine Vernet
#rstats 📦 {ergm} 4.12.0 is now on #CRAN. It's mostly bug fixes and incremental improvements, but the one you're most likely to notice is that most long-running routines now provide nice progress bars via {cli}, so you aren't left hanging. Other major changes at cran.r-project.org/package=ergm... .
R: NEWS file for the 'ergm' package
cran.r-project.org
February 18, 2026 at 12:35 AM
#rstats 📦 {ergm} 4.12.0 is now on #CRAN. It's mostly bug fixes and incremental improvements, but the one you're most likely to notice is that most long-running routines now provide nice progress bars via {cli}, so you aren't left hanging. Other major changes at cran.r-project.org/package=ergm... .
This is pretty cool!
In the large: Mortality in France, 1816-2016.
February 17, 2026 at 10:10 PM
This is pretty cool!
Reposted by Antoine Vernet
Surprise I wrote another quarto revealjs extension. Now giving you more fragments to use!
Easy to use fragments with conventional syntax. Works on headers, whole slides and on a letter-by-letter basis
github.com/EmilHvitfeld...
#quarto
Easy to use fragments with conventional syntax. Works on headers, whole slides and on a letter-by-letter basis
github.com/EmilHvitfeld...
#quarto
February 17, 2026 at 6:03 PM
Surprise I wrote another quarto revealjs extension. Now giving you more fragments to use!
Easy to use fragments with conventional syntax. Works on headers, whole slides and on a letter-by-letter basis
github.com/EmilHvitfeld...
#quarto
Easy to use fragments with conventional syntax. Works on headers, whole slides and on a letter-by-letter basis
github.com/EmilHvitfeld...
#quarto
Reposted by Antoine Vernet
Fact Check:
Nigel Farage claims the FT ignored the Mandelson-Epstein scandal
it was on the front page for SIX days
(long before he appears to have even noticed it)
Nigel Farage claims the FT ignored the Mandelson-Epstein scandal
it was on the front page for SIX days
(long before he appears to have even noticed it)
February 17, 2026 at 7:33 PM
Fact Check:
Nigel Farage claims the FT ignored the Mandelson-Epstein scandal
it was on the front page for SIX days
(long before he appears to have even noticed it)
Nigel Farage claims the FT ignored the Mandelson-Epstein scandal
it was on the front page for SIX days
(long before he appears to have even noticed it)
Tonight's experiment might only last 8 minutes...
@paulgp.com wasnt kidding, using the claude code container setup he shared (which is genuinely very helpful and useful), I blew through my token limit in… 8 minutes?!
February 17, 2026 at 8:20 PM
Tonight's experiment might only last 8 minutes...
Reposted by Antoine Vernet
In the old days, our hosts (Center for Open Science) used to provide regular data on uploads, downloads, and views. Now that they don't, I've been meaning to come up with a regular report mechanism. I don't know about APIs and JSON, but I am semi-fluent in Stata, so I tried using ChatGPT...
/1
/1
SocArXiv accepted 3,162 new papers in 2025, an increase of 20% over 2024. As we are now accepting only social science papers, and turning away new papers on technical AI topics, this seems to reflect an increase in (human) social science. Thanks for sharing, social scientists!
February 17, 2026 at 7:09 PM
In the old days, our hosts (Center for Open Science) used to provide regular data on uploads, downloads, and views. Now that they don't, I've been meaning to come up with a regular report mechanism. I don't know about APIs and JSON, but I am semi-fluent in Stata, so I tried using ChatGPT...
/1
/1
Reposted by Antoine Vernet
andreamatranga.github.io/CameraLensDa... I vibecoded an explorer for the canon and nikon SLR lens systems. I had made a stata version years ago but now it's interactive and online.
February 17, 2026 at 3:05 PM
andreamatranga.github.io/CameraLensDa... I vibecoded an explorer for the canon and nikon SLR lens systems. I had made a stata version years ago but now it's interactive and online.
Farage proving once again he is a vacuous joke of a politician
Nigel Farage lashes out at a FT journalist and refuses to answer the question.
Is this how Reform UK would ignore media questions if they got into government?
Is this how Reform UK would ignore media questions if they got into government?
February 17, 2026 at 5:45 PM
Farage proving once again he is a vacuous joke of a politician
Reposted by Antoine Vernet
My column in today's paper: on banning the under-16s from social media, the trade-offs and assumptions that the ban makes:
Perhaps we should all be banned from social media
Focusing only on under-16s obscures the lack of internet safeguards for everyone else
www.ft.com
February 17, 2026 at 10:21 AM
My column in today's paper: on banning the under-16s from social media, the trade-offs and assumptions that the ban makes:
I don't know, this seems like the kind of idea your internal processes should have killed. Several times
Meta has patented AI that can run a dead person's account, continuing to post and chat on their behalf
It can message and video call by replicating a user's online behavior using their past data
It can message and video call by replicating a user's online behavior using their past data
February 17, 2026 at 3:49 PM
I don't know, this seems like the kind of idea your internal processes should have killed. Several times
Sometimes I wonder: do these people just have access to nicer psychedelics?
February 17, 2026 at 3:44 PM
Sometimes I wonder: do these people just have access to nicer psychedelics?
Encouraging 🤪
Giving Braverman the education brief is quite a choice for Reform. I was on an education commission with her once before she went completely mad and she knew absolutely nothing at all.
February 17, 2026 at 3:11 PM
Encouraging 🤪
Reposted by Antoine Vernet
Reposted by Antoine Vernet
Once thronged with congregants; now sadly out of fashion as everyone attends the Church of the Collider instead.
February 17, 2026 at 1:22 PM
Once thronged with congregants; now sadly out of fashion as everyone attends the Church of the Collider instead.
This 👇
The drunk uncle theory.
You don’t argue with the casually homophobic uncle at Thanksgiving dinner to change his mind; you argue so that the closeted cousin at the kids table knows there’s safe people and better possibilities out there
You don’t argue with the casually homophobic uncle at Thanksgiving dinner to change his mind; you argue so that the closeted cousin at the kids table knows there’s safe people and better possibilities out there
agree with this (hah) but also think a particular mistake the left made for a long time online, and still makes to an extent, is failing to understand that the person whose mind you may actually change is the one reading the argument you're having, not the one you're arguing with
February 17, 2026 at 9:37 AM
This 👇
One of my favourite images
Everything is awful but there's a chance we will get this image on the cover of a special issue about statistical workflow. Really the best advertising decision the journal could make. I hope they go for it
February 17, 2026 at 9:34 AM
One of my favourite images
Reposted by Antoine Vernet
Everything is awful but there's a chance we will get this image on the cover of a special issue about statistical workflow. Really the best advertising decision the journal could make. I hope they go for it
February 17, 2026 at 8:46 AM
Everything is awful but there's a chance we will get this image on the cover of a special issue about statistical workflow. Really the best advertising decision the journal could make. I hope they go for it
Reposted by Antoine Vernet
Problem about the loneliness epidemic is, it's everywhere except in representative survey data. Let's look at where the claim comes from. 1/
Case in point, @pengzell.bsky.social just sent me 5 papers proving that there’s not actually any evidence for a loneliness epidemic. My mind is highly changed
February 17, 2026 at 7:13 AM
Problem about the loneliness epidemic is, it's everywhere except in representative survey data. Let's look at where the claim comes from. 1/
Reposted by Antoine Vernet
So yeah, already fourth iteration of my Data Analysis with AI course. gabors-data-analysis.com/ai-course/ All open source. I have an intro overview of AI and its use in data. gabors-data-analysis.com/courses/da-w...
February 16, 2026 at 10:32 PM
So yeah, already fourth iteration of my Data Analysis with AI course. gabors-data-analysis.com/ai-course/ All open source. I have an intro overview of AI and its use in data. gabors-data-analysis.com/courses/da-w...
Reposted by Antoine Vernet
I don't think I've ever seen gears shift this fast before.
February 15, 2026 at 9:50 PM
I don't think I've ever seen gears shift this fast before.
Reposted by Antoine Vernet
HELLO I WAS A QUEER TEEN IN THE 1990s AND THE ARGUMENTS AGAINST SOCIAL TRANSITION NOW ARE INDISTINGUISHABLE FROM THE ARGUMENTS AGAINST "LETTING" US BE (at the time, for me) GAY OR LESBIAN THEN.
They all rest on the belief that it's inherently better to be straight (then)/cis (now) and that's BS.
They all rest on the belief that it's inherently better to be straight (then)/cis (now) and that's BS.
Hilary Cass here arguing for children to be prevented from being known by a preferred name or getting a haircut in case it encourages them to continue later in life to transition.
Hilary Cass acts like gender variant children should have the transness hazed out of them.
Hilary Cass acts like gender variant children should have the transness hazed out of them.
February 16, 2026 at 6:32 PM
HELLO I WAS A QUEER TEEN IN THE 1990s AND THE ARGUMENTS AGAINST SOCIAL TRANSITION NOW ARE INDISTINGUISHABLE FROM THE ARGUMENTS AGAINST "LETTING" US BE (at the time, for me) GAY OR LESBIAN THEN.
They all rest on the belief that it's inherently better to be straight (then)/cis (now) and that's BS.
They all rest on the belief that it's inherently better to be straight (then)/cis (now) and that's BS.
Reposted by Antoine Vernet
Feb. update to the LLM+R guide 💪
8 new packages including:
code review, predictive modeling, speech-to-text, text-to-speech, HuggingFace integration, Gemini CLI companion, a CLI coding agent written in R🤯 ,and more!
available in English 🇱🇷 and Spanish 🇲🇽
luisdva.github.io/llmsr-book/
#rstats
8 new packages including:
code review, predictive modeling, speech-to-text, text-to-speech, HuggingFace integration, Gemini CLI companion, a CLI coding agent written in R🤯 ,and more!
available in English 🇱🇷 and Spanish 🇲🇽
luisdva.github.io/llmsr-book/
#rstats
February 16, 2026 at 5:30 PM
Feb. update to the LLM+R guide 💪
8 new packages including:
code review, predictive modeling, speech-to-text, text-to-speech, HuggingFace integration, Gemini CLI companion, a CLI coding agent written in R🤯 ,and more!
available in English 🇱🇷 and Spanish 🇲🇽
luisdva.github.io/llmsr-book/
#rstats
8 new packages including:
code review, predictive modeling, speech-to-text, text-to-speech, HuggingFace integration, Gemini CLI companion, a CLI coding agent written in R🤯 ,and more!
available in English 🇱🇷 and Spanish 🇲🇽
luisdva.github.io/llmsr-book/
#rstats
Reposted by Antoine Vernet
With bookdown.org being decommissioned, I've been working on converting #rstats books from #bookdown to #quartopub. After spending a lot of time with Claude to get it right for my two books, here's a repo that might help you do the same (especially with Claude Code):
github.com/topepo/bookd...
github.com/topepo/bookd...
February 16, 2026 at 2:55 PM
With bookdown.org being decommissioned, I've been working on converting #rstats books from #bookdown to #quartopub. After spending a lot of time with Claude to get it right for my two books, here's a repo that might help you do the same (especially with Claude Code):
github.com/topepo/bookd...
github.com/topepo/bookd...
I think this is true of a lot of people: some of their needs can be served with local, smaller models. But I think the question of what to send to which model is not trivial. I often end up using a SOTA model because given my low (-ish) usage, 1/2
These days most of my LLM tokens are spent on laptop, with open-weight models.
My situation is atypical: I have a lot of simple textual summary-like tasks that don’t need SOTA.
But I’m sharing just to say: SOTA can eventually overshoot need for some (small but growing) fraction of tasks. +
My situation is atypical: I have a lot of simple textual summary-like tasks that don’t need SOTA.
But I’m sharing just to say: SOTA can eventually overshoot need for some (small but growing) fraction of tasks. +
February 16, 2026 at 3:30 PM
I think this is true of a lot of people: some of their needs can be served with local, smaller models. But I think the question of what to send to which model is not trivial. I often end up using a SOTA model because given my low (-ish) usage, 1/2