The 252nd Block: "The truth is way more depressing"
Still not looking up
This week…
Your reading time is about 5 minutes. Let’s start.
Every day I am slowly trying to convince myself that I should watch Don’t Look Up (2021). I understand this film is in my wheelhouse and is even more relevant now more than ever. I have refused to watch anything with Leonardo DiCaprio for various reasons, but especially for his involvement in the 1MDB scandal (I am not THAT patriotic, but I will never forget). I don’t normally have particularly strong feelings about celebrities I don’t know personally, but for some reason this guy really annoys me. If there is a malicious way I would use AI, it would be to replace him with someone else in the film.
Anyway, I am writing my final assignment for my public law module due next week. It’s about human rights and unlawful detention. My brain is cooked, as the kids would say.
Your Wikipedia this week: Shopkeeper’s privilege
And now, a selection of top stories on my radar, a few personal recommendations, and the chart of the week.
ICYMI: The Previous Block was about democracy in Iran, Ukraine, Germany, and AI policies in the African and European countries. FWIW:
AI intensifies South China Sea disputes in the Philippines by Nuurrianti Jalli and Mangel Martinez for Fulcrum.
Generative AI, democracy and human rights by David Evan Harris and Aaron Shull for CIGI.
CORRECTION NOTICE: None notified.ON THE OLIGARCHY
Why the future of democracy depends on controlling illicit finance
Oliver Bullough for Coda:
The core of Vladimir Putin-style politics is what he understands winning an argument to look like. When his opponents are too scared, confused, exhausted, or dead to continue, he thinks he’s won. Sometimes he has: murdering anyone who disagreed with him in Chechnya, shattering an entire city, plus driving out hundreds of thousands of people, did indeed pacify that poor, beautiful place, though it did not work so well as a strategy in Syria.
But here’s why the truth is so troubling to oligarchs, and why Donald Trump unleashed his inner troll when Volodymyr Zelensky said some anodyne but true things, because, no matter how loud you shout, no matter how many people you imprison or murder, two plus two always equals four. And “if that is granted, all else follows.”
Eat the rich. Loosely linked:
Musk and Milei’s chainsaw bromance by Ana Prieto for Coda.
The anatomy of oligarchy by Matt Hatfield for Open Media.
Understanding money is key to grappling with power. Business journalism isn’t set up for that by Sara Silver for CJR.
ON GEOPOLITICS
How digital maps display disputed names
Viola Zhou and Khadija Alam for Rest of World:
After U.S. President Donald Trump issued an executive order renaming the Gulf of Mexico as Gulf of America in January, some digital maps quickly made the change, at least for American users.
“People using Maps in the U.S. will see ‘Gulf of America,’ and people in Mexico will see ‘Gulf of Mexico.’ Everyone else will see both names,” Google said in a blog post.
It wasn’t the first time that mapping companies have had to update a name. Around the world, many places have had multiple names at different times due to historical reasons or territorial disputes. Governments often pressure map companies, airlines, media outlets, and even streaming firms to adopt their preferred names.
Nine dash line, the West Philippine Sea, and others. Loosely linked:
The land disputes facing African Americans in Ghana by Philip Teye Agbove for New Lines Magazine.
Google removing ‘state’ designation from Canadian government buildings. Microsoft following suit by Andrew Kurjata for CBC.
German election rekindles old divisions by Hanne Cokelaere and Lucia Mackenzie for Politico.eu.
AI IN ACADEMIA
AIs flunk language test that takes grammar out of the equation
Rutvik Desai (University of South Carolina) for The Conversation:
The results suggest that large language models do not have the same sense-making capabilities as human beings. It is worth noting that our test relies on a subjective task, where the gold standard is ratings given by people. There is no objectively right answer, unlike typical large language model evaluation benchmarks involving reasoning, planning or code generation.
The low performance was largely driven by the fact that large language models tended to overestimate the degree to which a noun-noun pair qualified as meaningful. They made sense of things that should not make much sense. In a manner of speaking, the models were being too creative. One possible explanation is that the low-meaningfulness word pairs could make sense in some context. A beach covered with balls could be called a “ball beach.” But there is no common usage of this noun-noun combination among English speakers.
Is this like comparable to Chinese room? Loosely linked:
UK universities warned to ‘stress-test’ assessments as 92% of students use AI by Sally Weale for The Guardian.
We gave an AI a Rorschach test. What it saw in the inkblots offers a window into the human mind by Chris Stokel-Walker for BBC.
How to use NotebookLM, Google's research companion by Kaycee Hill for Tom’s Guide.
Other curious links, including en español et français
LONG READ | What’s the meaning of the objects an analyst puts on display by Anna Parker (University of London) for Aeon.
INFOGRAPHIC | The lives lost in Gaza: A closer look at those killed in the conflict so far by Sudev Kiyada, Simon Scarr and Nidal al-Mughrabi for Reuters.
PHOTO ESSAY | ‘We’ve lost some parts of nature for ever’: Ukraine war’s impact on environment by Luke Harding with photos by Alessio Mamo for The Guardian.
Revealed: at least 25 UK ‘spy cops’ had sex with deceived members of public by Rob Evans for The Guardian.
A journey through the hyper-political world of microchips by Helen Sullivan for The Guardian.
Filtros de belleza y su impacto en la salud mental: “No sé cómo antes me veía bien al hacerme fotos” por Marta Rey en RTVE.
El gran misterio detrás de los fallos de ChatGPT y el resto de inteligencias artificiales por Jesús Díaz en El Confidencial.
Neuroderechos: qué dice la neurociencia por Antonello Novelli Ciotti (Universidad de Oviedo) en The Conversation.
Letterboxd, Goodreads : pourquoi les applications culturelles rivalisent avec les réseaux sociaux ? par Marion Dos Santos Clara dans L’ADN.
Le cinéma boude-t-il le climat ? par Valérie Boisclair (texte) et Josselin Pfeuffer (illustration) dans Radio-Canada.
Anomalies visuelles, recherche inversée... Cinq conseils pour reconnaître une image générée par intelligence artificielle par Linh-Lan Dao dans Franceinfo.
What I read, listen, and watch
I’m reading Standoff (2021) by Bruce McIvor about why reconciliation fails Indigenous people. It’s quite history- and law-heavy, which would usually affect my overall comprehension, but I think two semesters of law school has improved my ability to grasp some of these political and legal concepts.
I’m listening to a Don’t Call Me Resilient episode on AI-generated influencers and whether it is a form of cultural exploitation.
I’m watching Companion (2025) a horror comedy sci-fi about our relationship with AI. Edinburgh Napier University’s Sarah Artt wrote a piece about it for The Conversation last month.
Chart of the week
The EIU launched its 2024 Democracy Index this week, showing a trend of global democratic decline and increasing authoritarianism. Axios’ Avery Lotz visualised the result in an interactive map here.