{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreih73lpjw47voyz2tyfnmd7egua2bpwoituuynf3tss2i4f7lhx7i4",
    "uri": "at://did:plc:zyvv64s26uy7h2rwhhdq5e5f/app.bsky.feed.post/3mjclb76z6dk2"
  },
  "path": "/2026/04/02/bookmarks/",
  "publishedAt": "2026-04-02T04:00:00.000Z",
  "site": "https://inkdroid.org",
  "tags": [
    "Review: Measuring AI Ability to Complete Long Software Tasks",
    "RADIO CAUSE\nCOMMUNE 93.1 FM • PARIS",
    "Pourquoi je n’utilise pas l’IA",
    "London Book\nTrades Database",
    "Wikipedia Bans AI-Generated Content",
    "Trump administration requests Stanford Medical School admissions data,\nclaiming racial discrimination",
    "Discounted Cumulative Gain",
    "axios Compromised on npm - Malicious Versions Drop Remote Access Trojan",
    "webweigh",
    "Phosphor Icons",
    "Departures (2008 film)",
    "Infrastructure Landlords: The Rentier Capitalism of Commercial Academic\nPublishers",
    "AI got the blame for the Iran school bombing. The truth is far more\nworrying",
    "Guibo",
    "The\nHuman Line Project",
    "Marriage over, €100,000 down the drain: the AI users whose lives were\nwrecked by delusion"
  ],
  "textContent": "These are some things I’ve wandered across on the web this week.\n\n##  🔖 \nReview: Measuring AI Ability to Complete Long Software Tasks \n\nMeasuring AI Ability to Complete Long Software Tasks, a paper by dozens of authors working at Model Evaluation & Threat Research (METR). They define the “time horizon” metric and show that LLMs’ time horizons have been doubling every seven months, and this growth might have recently accelerated.\n\n##  🔖  RADIO CAUSE\nCOMMUNE 93.1 FM • PARIS \n\nRadio Cause Commune est une radio associative parisienne qui diffuse depuis novembre 2017 sur 93.1 FM. 40 bénévoles, zéro publicité, un budget de 60 000€ annuel : nous maintenons une stricte indépendance éditoriale. Nous défendons les logiciels libres, l’indépendance des médias et créons des outils techniques innovants pour la radiophonie libre\n\n##  🔖 \nPourquoi je n’utilise pas l’IA \n\nL’IA me gonfle. Profondément. Enfin, surtout l’IA générative (tu sais, les LLM), parce que je peux concevoir une certaine utilité à certains types d’IA. La reconnaissance vocale, par exemple.\n\nPassons un peu en revue mes raisons de ne pas utiliser l’IA.\n\n##  🔖  London Book\nTrades Database \n\nThe Bibliographical Society has just launched a redesigned version of the London Book Trades Database (https://lbt.bibsoc.org.uk/).\n\nThe original LBT database was the work of the late Michael Turner at the Bodleian Library, assisted by a number of collaborators, drawing particularly on the archival resources of the Stationers’ Company. A web version of the database was created in 2009 which eventually ran on servers at the Bodleian until it was closed down in 2024 as its software was long past its expiry date.\n\nThe Bibliographical Society has taken steps to revive the project, this time as a read-only MediaWiki resource based on a new extraction of the data from the original database created by Michael Turner and a radical redesign of the contents and interface (I led this work). This new version, known as LBT Version 2, does not yet contain all the original data, but the people, events, titles, and relationships make it immediately useful. We envisage two or three updates in the coming months as more contents are retrieved and restructured. The new web site has explanatory pages with a full history of the project and its new technical implementation.\n\nIn addition to all the famous names of the book trade up to the mid-nineteenth century, entries offer information for more minor figures including family members and apprentices. There are entries for nearly 35,000 people, presenting detailed accounts of the person’s interaction with the Stationers’ Company and data from published sources.\n\n##  🔖 \nWikipedia Bans AI-Generated Content \n\nAfter months of heated debate and previous attempts to restrict the use of large language models on Wikipedia, on March 20 volunteer editors accepted a new policy that prohibits using them to create articles for the online encyclopedia.\n\n“Text generated by large language models (LLMs) often violates several of Wikipedia’s core content policies,” Wikipedia’s new policy states. “For this reason, the use of LLMs to generate or rewrite article content is prohibited, save for the exceptions given below.”\n\n##  🔖 \nTrump administration requests Stanford Medical School admissions data,\nclaiming racial discrimination \n\nThe Trump Administration opened investigations into admissions policies at the medical schools of Stanford University, Ohio State University (OSU) and the University of California, San Diego (UCSD) on March 25, noting possible race discrimination.\n\nIn letters sent to the three schools, the Department of Justice (DOJ) requested data on the last seven years of admitted classes at the medical schools, threatening to withhold federal funding if the schools do not comply by turning over the data requested by April 24. The investigation is part of a larger crackdown on higher education, as the DOJ has launched dozens of investigations into universities during Trump’s second term.\n\n##  🔖 \nDiscounted Cumulative Gain \n\nDiscounted cumulative gain (DCG) is a measure of ranking quality in information retrieval. It is often normalized so that it is comparable across queries, giving Normalized DCG (nDCG or NDCG). NDCG is often used to measure effectiveness of search engine algorithms and related applications. Using a graded relevance scale of documents in a search-engine result set, DCG sums the usefulness, or gain, of the results discounted by their position in the result list.[1] NDCG is DCG normalized by the maximum possible DCG of the result set when ranked from highest to lowest gain, thus adjusting for the different numbers of relevant results for different queries.\n\n##  🔖 \naxios Compromised on npm - Malicious Versions Drop Remote Access Trojan\n\n\naxios is the most popular JavaScript HTTP client library with over 100 million weekly downloads. On March 30, 2026, StepSecurity identified two malicious versions of the widely used axios HTTP client library published to npm: axios@1.14.1 and axios@0.30.4. The malicious versions inject a new dependency, plain-crypto-js@4.2.1, which is never imported anywhere in the axios source code. Its sole purpose is to execute a postinstall script that acts as a cross platform remote access trojan (RAT) dropper, targeting macOS, Windows, and Linux. The dropper contacts a live command and control server and delivers platform specific second stage payloads. After execution, the malware deletes itself and replaces its own package.json with a clean version to evade forensic detection.\n\n##  🔖 \nwebweigh \n\nA rust CLI that calculates the file size of a web page when loaded with all external resources.\n\n##  🔖  Phosphor Icons\n\n\nPhosphor is a flexible icon family for interfaces, diagrams, presentations — whatever, really.\n\n##  🔖 \nDepartures (2008 film) \n\nepartures (Japanese: おくりびと, Hepburn: Okuribito; “one who sends off”) is a 2008 Japanese black comedy drama film directed by Yōjirō Takita and starring Masahiro Motoki, Ryōko Hirosue, and Tsutomu Yamazaki. The film follows a young man who returns to his hometown after a failed career as a cellist and stumbles across work as a nōkanshi—a traditional Japanese ritual mortician. He is subjected to prejudice from those around him, including from his wife, because of strong social taboos against people who deal with death. Eventually he repairs these interpersonal connections through the beauty and dignity of his work.\n\n##  🔖 \nInfrastructure Landlords: The Rentier Capitalism of Commercial Academic\nPublishers \n\nIf you want to understand where the commercial parts of scholarly communications may be heading, you need to look beyond policy documents, conference panels, or public-facing strategy statements. You should look at what large commercial actors say when speaking to investors. Earnings calls are one of the places where that language becomes especially revealing: less concerned with sector ideals than with growth, market opportunity, competitive position, and what will ultimately generate value for shareholders. For this reason, it can be worthwhile to review earnings calls and investor presentations, as these are often overlooked when discussing OA policy and sectoral movements.\n\n##  🔖 \nAI got the blame for the Iran school bombing. The truth is far more\nworrying \n\nSomeone decided to compress the kill chain. Someone decided that deliberation was latency. Someone decided to build a system that produces 1,000 targeting decisions an hour and call them high-quality. Someone decided to start this war. Several hundred people are sitting on Capitol Hill, refusing to stop it. Calling it an “AI problem” gives those decisions, and those people, a place to hide.\n\n##  🔖  Guibo \n\nGUIBo is a desktop GUI for operators and developers who run Kubo (the IPFS daemon in Go). It drives your node through Kubo’s HTTP RPC API so you can work with pins, UnixFS content, IPNS, remote pinning, gateways, and network or repo diagnostics without living in the terminal.\n\n##  🔖  The\nHuman Line Project \n\nAt The Human Line, we are committed to ensuring that AI technologies, like chatbots, are developed and deployed with the human element at their core. LLMs are powerful tools, and with Ethical design, users can gain new skills and knowledge while remaining emotionally intact.\n\n##  🔖 \nMarriage over, €100,000 down the drain: the AI users whose lives were\nwrecked by delusion \n\nTech-related delusions, whether they involve train travel, radio transmitters or 5G masts, have been around for centuries, Morrin says. “What’s different is that we’re now arguably entering an age in which people aren’t having delusions about technology, but having delusions with technology. What’s new is this co-construction, where technology is an active participant. AI chatbots can co-create these delusional beliefs.”",
  "title": "Weekly Bookmarks"
}