Daily Top Stories
1.
Meta Segment Anything Model 3 (ai.meta.com)
2.
The patent office is about to make bad patents untouchable (eff.org)
3.
Gaming on Linux has never been more approachable (theverge.com)
4.
AI is a front for consolidation of resources and power (chrbutler.com)
5.
Building more with GPT-5.1-Codex-Max (openai.com)
6.
The Death of Arduino? (linkedin.com)
7.
Loose wire leads to blackout, contact with Francis Scott Key bridge (ntsb.gov)
8.
Researchers discover security vulnerability in WhatsApp (univie.ac.at)
9.
Cognitive and mental health correlates of short-form video use (psycnet.apa.org)
10.
Precise geolocation via Wi-Fi Positioning System (amoses.dev)
11.
Microsoft AI CEO pushes back against critics after recent Windows AI backlash (windowscentral.com)
12.
PHP 8.5 (stitcher.io)
13.
Static Web Hosting on the Intel N150 (it-notes.dragas.net)
14.
Verifying your Matrix devices is becoming mandatory (element.io)
15.
Basalt Woven Textile (materialdistrict.com)
16.
How Slide Rules Work (amenzwa.github.io)
17.
Interactive World History Atlas Since 3000 BC (geacron.com)
18.
Launch HN: Mosaic (YC W25) – Agentic Video Editing (mosaic.so)
19.
Gov. Abbott's office redacts pages of emails about Elon Musk (kut.org)
20.
40 years ago, Calvin and Hobbes' burst onto the page (npr.org)
21.
The lost cause of the Lisp machines (tfeb.org)
22.
Measuring political bias in Claude (anthropic.com)
23.
Tailscale Down (status.tailscale.com)
24.
Measuring the impact of AI scams on the elderly (simonlermen.substack.com)
25.
Netherlands returns control of Nexperia to Chinese owner (bloomberg.com)
26.
LLMs are bullshitters. But that doesn't mean they're not useful (blog.kagi.com)
27.
Adversarial Poetry as a Universal Single-Turn Jailbreak Mechanism in LLMs (arxiv.org)
28.
CornHub (cornhub.website)
29.
Emoji evidence errors don’t undo a murder conviction (blog.ericgoldman.org)
30.
I am just sooo sick of AI prediction content, let's kill it already (verdikapuku.com)