Alignment pretraining: AI discourse creates self-fulfilling (mis)alignment
https://arxiv.org/abs/2601.10160
https://arxiv.org/abs/2601.10160
arXiv.org
Alignment Pretraining: AI Discourse Causes Self-Fulfilling (Mis)alignment
Pretraining corpora contain extensive discourse about AI systems, yet the causal influence of this discourse on downstream alignment remains poorly understood. If prevailing descriptions of AI...
Mexican government breached by solo user with Claude, 150 GB exfiltrated
https://konstantintkachuk.com/writing/the-floor-doesnt-exist/
https://konstantintkachuk.com/writing/the-floor-doesnt-exist/
Konstantintkachuk
The Floor Doesn't Exist
AI did not invent new attacks. It billed the old ones monthly. Crypto is the only place we can count it, and the early signals are in.
Anthropic co-founder to present AI encyclical alongside Pope Leo XIV
https://www.vaticannews.va/en/pope/news/2026-05/pope-leo-xiv-first-encyclical-magnifica-humanitas.html
https://www.vaticannews.va/en/pope/news/2026-05/pope-leo-xiv-first-encyclical-magnifica-humanitas.html
www.vaticannews.va
Pope Leo XIV’s first encyclical Magnifica humanitas to be published May 25 - Vatican News
Pope Leo XIV’s first encyclical, Magnifica humanitas, on safeguarding the human person in the time of artificial intelligence, will be released on May ...
What political censorship looks like inside an LLM's weights (Qwen 3.5)
https://vas-blog.pages.dev/qwen-censorship/
https://vas-blog.pages.dev/qwen-censorship/
Microsoft surprises with its first server Linux distribution: Azure Linux 4.0
https://www.zdnet.com/article/microsoft-releases-its-first-server-linux-distribution-azure-linux-4-0/
https://www.zdnet.com/article/microsoft-releases-its-first-server-linux-distribution-azure-linux-4-0/
ZDNET
Microsoft surprises with its first server Linux distribution: Azure Linux 4.0
You'll be able to run this Linux distro on both Azure and your desktop using Windows Subsystem for Linux. Here's what we know about it so far.
The American Rebellion Against AI Is Gaining Steam
https://www.wsj.com/tech/ai/the-american-rebellion-against-ai-is-gaining-steam-94b72529
https://www.wsj.com/tech/ai/the-american-rebellion-against-ai-is-gaining-steam-94b72529
The Wall Street Journal
The American Rebellion Against AI Is Gaining Steam
Booed commencement speakers, blocked data centers, plummeting poll numbers: The fast-growing industry has a faster-growing crisis.
Mini Shai-Hulud Strikes Again: 314 npm Packages Compromised
https://safedep.io/mini-shai-hulud-strikes-again-314-npm-packages-compromised/
https://safedep.io/mini-shai-hulud-strikes-again-314-npm-packages-compromised/
SafeDep - Real-time Open Source Software Supply Chain Security
Mini Shai-Hulud Strikes Again: 317 npm Packages Compromised
A compromised npm maintainer account published 637 malicious versions across 317 packages including size-sensor, echarts-for-react, timeago.js, and hundreds of @antv scoped packages, affecting 15M+ monthly downloads.
Show HN: Id-agent – Token efficient UUID alternative for AI agents
https://github.com/vostride/id-agent
https://github.com/vostride/id-agent
GitHub
GitHub - vostride/id-agent: Token efficient IDs for AI agents. UUID alternative for the agentic era
Token efficient IDs for AI agents. UUID alternative for the agentic era - vostride/id-agent
New accessibility features powered by Apple Intelligence
https://www.apple.com/newsroom/2026/05/apple-unveils-new-accessibility-features-and-updates-with-apple-intelligence/
https://www.apple.com/newsroom/2026/05/apple-unveils-new-accessibility-features-and-updates-with-apple-intelligence/
Apple Newsroom
Apple unveils new accessibility features, and updates with Apple Intelligence
Apple announced major accessibility updates powered by Apple Intelligence, including new capabilities for VoiceOver, Magnifier, and Voice Control.
Anthropic Is Preparing for IPO and We Should Be Worried
https://www.vincentschmalbach.com/anthropic-ipo-developers-should-be-worried-v2/
https://www.vincentschmalbach.com/anthropic-ipo-developers-should-be-worried-v2/
Vincent Schmalbach
Anthropic is Preparing for IPO and We Should Be Worried
Anthropic is starting to act like a company preparing for public markets: protecting margins, tightening access. The gap between the brand and the behavior is now too obvious to ignore.
Launch HN: Superlog (YC P26) – Observability that installs itself and fixes bugs
https://superlog.sh/
https://superlog.sh/
'Comically bad' datasets used to train clinical models for stroke and diabetes
https://retractionwatch.com/2026/05/18/kaggle-dataset-clinical-models-stroke-diabetes/
https://retractionwatch.com/2026/05/18/kaggle-dataset-clinical-models-stroke-diabetes/
Retraction Watch
‘Comically bad’ datasets used to train clinical models for stroke and diabetes
A dataset on Kaggle purportedly showing people who have had a stroke includes images of Sylvester Stallone from Rambo and other celebrities. Source Scrolling through an online image dataset, Adrian…