Laurence Dierickx
@ohmyshambles.bsky.social
Researcher in information/disinformation #AI technology with a blend of STS, ethics and journalism studies 🇧🇪 🇳🇴 https://ohmybox.info/about
Pinned
New publication alert! From bytes to bylines - A history of AI in journalism practices
www.taylorfrancis.com/chapters/edi...
www.taylorfrancis.com/chapters/edi...
Reposted by Laurence Dierickx
Ben was the lead data journalism developer at the LATimes for many years. He’s open sourcing his new news apps course materials!
📚 I'm pleased to release my syllabus and course materials for "Coding the News," the class I'm leading this semester for the Craig Newmark Graduate School of Journalism at CUNY.
February 10, 2026 at 3:38 PM
Ben was the lead data journalism developer at the LATimes for many years. He’s open sourcing his new news apps course materials!
Oui, c'est effrayant et ça l'est d'autant plus que ce rédacteur en chef intervient régulièrement dans certaines écoles de journalisme. Comme quoi la route vers l'éducation à l'IA est encore longue et semée d'embûches !
February 10, 2026 at 9:59 AM
Oui, c'est effrayant et ça l'est d'autant plus que ce rédacteur en chef intervient régulièrement dans certaines écoles de journalisme. Comme quoi la route vers l'éducation à l'IA est encore longue et semée d'embûches !
Cette déclaration suggère l’une de ces deux choses :
– une incompréhension du fonctionnement des LLMs, ou
– une inquiétante complaisance à l’égard d’une génération probabiliste traitée comme un jugement éditorial.
Le même trouve que les "facultés" de l'IA sont "stupéfiantes". Sans voix. (2/2)
– une incompréhension du fonctionnement des LLMs, ou
– une inquiétante complaisance à l’égard d’une génération probabiliste traitée comme un jugement éditorial.
Le même trouve que les "facultés" de l'IA sont "stupéfiantes". Sans voix. (2/2)
February 10, 2026 at 9:00 AM
Cette déclaration suggère l’une de ces deux choses :
– une incompréhension du fonctionnement des LLMs, ou
– une inquiétante complaisance à l’égard d’une génération probabiliste traitée comme un jugement éditorial.
Le même trouve que les "facultés" de l'IA sont "stupéfiantes". Sans voix. (2/2)
– une incompréhension du fonctionnement des LLMs, ou
– une inquiétante complaisance à l’égard d’une génération probabiliste traitée comme un jugement éditorial.
Le même trouve que les "facultés" de l'IA sont "stupéfiantes". Sans voix. (2/2)
En Belgique, le rédacteur en chef de ce qui est considéré un quotidien de qualité reconnaît, en toute transparence, demander à l'IA générative "ce qu'il pense d'un projet de texte. Je le soumets à l'IA comme je le soumettrais à un journaliste de la rédaction." (1/2)
February 10, 2026 at 9:00 AM
En Belgique, le rédacteur en chef de ce qui est considéré un quotidien de qualité reconnaît, en toute transparence, demander à l'IA générative "ce qu'il pense d'un projet de texte. Je le soumets à l'IA comme je le soumettrais à un journaliste de la rédaction." (1/2)
1) a misunderstanding of how large language models actually work, or 2) a troubling comfort with probabilistic text generation being treated as editorial judgment. Either way, that’s pretty scary. (2/2)
February 10, 2026 at 8:55 AM
1) a misunderstanding of how large language models actually work, or 2) a troubling comfort with probabilistic text generation being treated as editorial judgment. Either way, that’s pretty scary. (2/2)
In Fr. Belgium, the editor-in-chief of a newspaper widely regarded as a quality publication admits using generative AI to ask: “What it thinks of a draft text. I submit it to AI the same way I would submit it to a journalist in the newsroom.” This statement suggests one of two things: (1/2)
February 10, 2026 at 8:55 AM
In Fr. Belgium, the editor-in-chief of a newspaper widely regarded as a quality publication admits using generative AI to ask: “What it thinks of a draft text. I submit it to AI the same way I would submit it to a journalist in the newsroom.” This statement suggests one of two things: (1/2)
Reposted by Laurence Dierickx
Claims of “intelligence” focus on function: if AI behaves “as if” it has a mind, it’s “intelligent.” But behavior is useless when models are designed to pass behavioral tests. Instead, the AGI label achieves the opposite goal: it suggests a human-like experience of being an LLM. ⬇️
The most important question about AI isn't determining whether the machine thinks — it's whether interacting with it displaces the conditions under which we think, writes Eryk Salvaggio. www.techpolicy.press/the-illusion...
The Illusion of AGI, or What Language Models Can Do Without Thought
It is not simple stubbornness that LLMs are not “intelligent,” much less a form of “general” intelligence, writes Eryk Salvaggio.
www.techpolicy.press
February 9, 2026 at 4:46 PM
Claims of “intelligence” focus on function: if AI behaves “as if” it has a mind, it’s “intelligent.” But behavior is useless when models are designed to pass behavioral tests. Instead, the AGI label achieves the opposite goal: it suggests a human-like experience of being an LLM. ⬇️
Reposted by Laurence Dierickx
I have been a little obsessed by the idea of a printed edition of the full parameters of an LLM. Never mind that it would take hundreds of years to do a single inference calculation by hand, it would be possible in theory, if you had the weights.
So I had Claude make it: weights-press.netlify.app
So I had Claude make it: weights-press.netlify.app
February 8, 2026 at 8:59 PM
I have been a little obsessed by the idea of a printed edition of the full parameters of an LLM. Never mind that it would take hundreds of years to do a single inference calculation by hand, it would be possible in theory, if you had the weights.
So I had Claude make it: weights-press.netlify.app
So I had Claude make it: weights-press.netlify.app
In 2016, filmmaker Oscar Sharp and technologist Ross Goodwin trained an AI on sci-fi scripts. They then had just 48 hours to film the bizarre screenplay that the AI generated, resulting in Sunspring. youtu.be/LY7x2Ihqjmc?...
Sunspring | A Sci-Fi Short Film Starring Thomas Middleditch
YouTube video by Ars Technica
youtu.be
February 6, 2026 at 4:55 PM
In 2016, filmmaker Oscar Sharp and technologist Ross Goodwin trained an AI on sci-fi scripts. They then had just 48 hours to film the bizarre screenplay that the AI generated, resulting in Sunspring. youtu.be/LY7x2Ihqjmc?...
Reposted by Laurence Dierickx
J'ai répondu aux questions de Damien Roulette (RTBF) à propos du "phénomène" Moltbook. Moltbook, le réseau social des IA : coup marketing ou début d’une dystopie technologique ? www.rtbf.be/article/molt...
Moltbook, le réseau social des IA : coup marketing ou début d’une dystopie technologique ? - RTBF Actus
Lancé mercredi 28 janvier, Moltbook a tous les apparats du forum Reddit mais il se présente comme un réseau...
www.rtbf.be
February 5, 2026 at 3:00 PM
J'ai répondu aux questions de Damien Roulette (RTBF) à propos du "phénomène" Moltbook. Moltbook, le réseau social des IA : coup marketing ou début d’une dystopie technologique ? www.rtbf.be/article/molt...
Washington Post cuts a third of its staff in a blow to a legendary news brand apnews.com/article/wash...
Washington Post cuts a third of its staff in a blow to a legendary news brand
The Washington Post has laid off one-third of its staff, eliminating its sports section, several foreign bureaus and its books coverage.
apnews.com
February 5, 2026 at 2:08 PM
Washington Post cuts a third of its staff in a blow to a legendary news brand apnews.com/article/wash...
Version originale en français : L'éthique journalistique est-elle mal en point ? ohmybox.info/carte-blanch...
Carte blanche : L’éthique journalistique est-elle mal en point ?
[caption id=attachment_5667 align=aligncenter width=820] Ces images générées par lIA illustrent les dérives de lusage des IA génératives ...
ohmybox.info
January 31, 2026 at 8:22 PM
Version originale en français : L'éthique journalistique est-elle mal en point ? ohmybox.info/carte-blanch...
Is journalism ethics in crisis? This keynote was presented at the French National Conference of Journalism Professions (CNJM), organised by the Journalism School of Tours on 29 January 2026. ohmybox.info/keynote-is-j...
[Keynote] Is journalism ethics in crisis?
Keynote presented at the National Conference of Journalism Professions (CNJM), organised at the journalism school of Tours, on January 29, 2026. 1. ...
ohmybox.info
January 31, 2026 at 8:20 PM
Is journalism ethics in crisis? This keynote was presented at the French National Conference of Journalism Professions (CNJM), organised by the Journalism School of Tours on 29 January 2026. ohmybox.info/keynote-is-j...
Inside Moltbook: The Social Network Where AI Agents Talk And Humans Just Watch www.forbes.com/sites/guneyy...
Moltbook AI Social Network: 1.4 Million Agents Build A Digital Society
A new platform, Moltbook, claims to host 1.4 million AI agents in a closed society. Discover why this 'hive mind' signals a dangerous reversal in human cognitive skills.
www.forbes.com
January 31, 2026 at 8:18 PM
Inside Moltbook: The Social Network Where AI Agents Talk And Humans Just Watch www.forbes.com/sites/guneyy...
There’s a social network for AI agents, and it’s getting weird www.theverge.com/ai-artificia...
There’s a social network for AI agents, and it’s getting weird
A viral post asks questions about consciousness.
www.theverge.com
January 31, 2026 at 8:17 PM
There’s a social network for AI agents, and it’s getting weird www.theverge.com/ai-artificia...
Google's new artificial intelligence (AI) search feature is facing criticism for providing erratic, inaccurate answers. www.bbc.com/news/article...
Google AI search tells users to glue pizza and eat rocks
Google has defended the answers given by AI Overview, describing them as "isolated examples".
www.bbc.com
January 31, 2026 at 7:25 AM
Google's new artificial intelligence (AI) search feature is facing criticism for providing erratic, inaccurate answers. www.bbc.com/news/article...
Je n'avais pas vu la première partie. Les moteurs de recherche "augmentés" vont rendre la vie impossible pour les recherches documentaires. Voir ici : www.rtbf.be/article/quan... et ici aussi ici.radio-canada.ca/nouvelle/218...
Quand le mode IA de Google valide une 'pluie de poissons', elle-même générée par IA - RTBF Actus
La vidéo montre des poissons qui tombent du ciel et atterrissent en masse sur les trottoirs d’une ville aux...
www.rtbf.be
January 31, 2026 at 7:22 AM
Je n'avais pas vu la première partie. Les moteurs de recherche "augmentés" vont rendre la vie impossible pour les recherches documentaires. Voir ici : www.rtbf.be/article/quan... et ici aussi ici.radio-canada.ca/nouvelle/218...
Reposted by Laurence Dierickx
Cherchant un article de JL Weissberg sur la fin de l'image comme preuve d'un événement passé, j'ai testé Gemini et obtenu un article inexistant dans une revue qui ne mentionne pas l'auteur.
January 29, 2026 at 3:19 PM
Cherchant un article de JL Weissberg sur la fin de l'image comme preuve d'un événement passé, j'ai testé Gemini et obtenu un article inexistant dans une revue qui ne mentionne pas l'auteur.
Merci pour la référence, excellent ! shs.cairn.info/article/HERM...
shs.cairn.info
January 29, 2026 at 3:44 PM
Merci pour la référence, excellent ! shs.cairn.info/article/HERM...
Reposted by Laurence Dierickx
🏛️ The AI-generated visual disinformation being spread by the White House regarding a recent protester arrest is damaging the already dwindling credibility of photos as proof.
Our lab director @claesdevreese.bsky.social has spoken with @rtl.nl and @france24.com about this: tinyurl.com/3y3fvbz9
Our lab director @claesdevreese.bsky.social has spoken with @rtl.nl and @france24.com about this: tinyurl.com/3y3fvbz9
White House spreads AI-manipulated image of protester arrest – AI, Media & Democracy Lab
www.aim4dem.nl
January 29, 2026 at 12:01 PM
🏛️ The AI-generated visual disinformation being spread by the White House regarding a recent protester arrest is damaging the already dwindling credibility of photos as proof.
Our lab director @claesdevreese.bsky.social has spoken with @rtl.nl and @france24.com about this: tinyurl.com/3y3fvbz9
Our lab director @claesdevreese.bsky.social has spoken with @rtl.nl and @france24.com about this: tinyurl.com/3y3fvbz9
Tracking AI-enabled Misinformation: 2,089 Undisclosed AI-Generated News Websites (and Counting) www.newsguardtech.com/special-repo...
Tracking AI-enabled Misinformation: Over 2000 Undisclosed AI-Generated News Websites (and Counting), Plus the Top False Narratives Generated by Artificial Intelligence Tools
www.newsguardtech.com
January 29, 2026 at 12:51 PM
Tracking AI-enabled Misinformation: 2,089 Undisclosed AI-Generated News Websites (and Counting) www.newsguardtech.com/special-repo...
More Articles Are Now Created by AI Than Humans #CNMJ graphite.io/five-percent...
More Articles Are Now Created by AI Than Humans
AI-generated content is as good or better than content written by humans. It is often hard to distinguish whether content is created by AI vs. a human. We seek to evaluate the prevalence of article co...
graphite.io
January 29, 2026 at 12:49 PM
More Articles Are Now Created by AI Than Humans #CNMJ graphite.io/five-percent...
Comprendre l'IA, c'est aussi pouvoir enquêter sur l'IA, voir le guide de l'UNESCO "Reporting on Artificial Intelligence" www.unesco.org/en/articles/...
Reporting on Artificial Intelligence
A Handbook for Journalism Educators
www.unesco.org
January 29, 2026 at 11:43 AM
Comprendre l'IA, c'est aussi pouvoir enquêter sur l'IA, voir le guide de l'UNESCO "Reporting on Artificial Intelligence" www.unesco.org/en/articles/...
Des pratiques éthique de l'IA, cela passe aussi par la compréhension de la technologie, cette plateforme explique assez simplement la complexité des modèles Transformer : poloclub.github.io/transformer-... #CNMJ
Transformer Explainer: LLM Transformer Model Visually Explained
An interactive visualization tool showing you how transformer models work in large language models (LLM) like GPT.
poloclub.github.io
January 29, 2026 at 11:36 AM
Des pratiques éthique de l'IA, cela passe aussi par la compréhension de la technologie, cette plateforme explique assez simplement la complexité des modèles Transformer : poloclub.github.io/transformer-... #CNMJ