CT scans of BYD car parts
BYD sold 4.6M vehicles in 2025, but Americans can't buy one. We CT scanned four components from its lineup to reveal what vertical integration looks like at the part level.
The latest blog posts to take off on Hacker News. 31,664 posts.
BYD sold 4.6M vehicles in 2025, but Americans can't buy one. We CT scanned four components from its lineup to reveal what vertical integration looks like at the part level.
Reading the screenshots, diagrams and tables in technical documentation for LLMs
Linux kernel experiments implementing IB verbs over consumer Thunderbolt/USB4 hardware to give cheap mini-PCs high-performance RDMA suitable for cluster inference.
coveillance.org
We’re extending Project Glasswing to approximately 150 new organizations in more than fifteen countries
Apple rejected WhisperPad under Guideline 2.4.5 for using the accessibility API to paste transcribed text. After an appeal that ended in another rejection, I split it into two versions — a compliant one on the App Store, and the original auto-paste version, direct.
We’ve always found strace useful but somewhat hard to work with. Its output is often inscrutable, it’s hard to follow subprocesses or threads, and if you wan...
So-called age verification for social media is spreading across the world, framed as an effort to create a safer internet for children. In reality, age verification lays the foundation for a fully government controlled internet.
OpenAI frontier models and Codex are now generally available on AWS, giving enterprises a new path to build with OpenAI through the AWS environments, controls, and procurement workflows they already use. Customers can get started with OpenAI on AWS and move faster from evaluation to production.
Anthropic has confidentially submitted a draft S-1 registration statement to the Securities and Exchange Commission
KDE is an open community of friendly people who want to create a world in which everyone has control over their digital life and enjoys freedom and privacy.
Physical AI systems must understand the real world before they can act within it. Robots, autonomous vehicles, and smart spaces need to understand what’s happening in their world…
The Town of Atherton's lawsuit against Caltrain electrification is the clearest case study we have of how a tiny, wealthy minority can hijack a…
ChatGPT for Google Sheets is vulnerable to data exfiltration and phishing overlay attacks that affect workbooks across the victim’s account after an indirect prompt injection in a single sheet.
Today we're releasing Bonsai Image 4B, compact 1-bit and ternary diffusion models built on FLUX.2 Klein 4B that run image generation on laptops and phones.
An investigation into fabricated citations and broken references in EY's "Points of Attack" cybersecurity report on loyalty system fraud.
Accenture has entered into an agreement to acquire Ookla, a global leader in network intelligence, competitive benchmarking and customer experience analytics.
A largely overlooked space between cells in women’s brains may hold the key to understanding memory loss tied to estrogen decline after menopause, reports a new preclinical Northwestern Medicine study.
TL;DR: MCP eats context, has low reliability, and overlaps with existing CLI/API.
DBOS recently argued that Postgres is all you need for durable execution: if you already trust your database, you do not need a separate orchestration tier. I a…
Explore Liquid AI’s LFM2.5-8B-A1B, an on-device Mixture of Experts model with 128K context, fast tool calling, and strong AI benchmarks.
An accidentally fast variable-length integer encoding
By Terts Diepraam Almost exactly one year ago, we announced Roto, a JIT-compiled embedded scripting language for Rust applications. A lot has happened since then that we'd like to tell you about! 💡Along with this post, we published Roto v0.11.0! You can check out the changelog
AI is doing to programming what framework-brain did to the frontend before. Deskilling, or just working at a higher level of abstraction?
Today, Kog AI launches a tech preview of the Kog Inference Engine (KIE): 3,000 output tokens/s per request on 8× AMD MI300X GPUs and 2,100 on 8× NVIDIA H200 (FP16, no speculative decoding). This preview runs a 2B model, with support for large third-party MoE models coming next at similar speeds.
New study by Emanuela Campisi (University of Catania) and Anita Slominska and Asli Ozyurek (Max Planck Institute for Psycholinguistics) reveals that Italian and Dutch adults adapt their hand gestures in remarkably similar ways when explaining new concepts to children.