Link Sharing

Risk of an unprecedented negative adjustment prompts plan to let UTC drift by as much as an hour

Porting Linux to Apple Silicon

posts and projects

ARC Prize scored GPT-6 Astra at 62.7% on ARC-AGI-3 with its own harness and 99.9% with OpenAI's, then revised five metrics after launch.

With 'GPT-6 Astra', AGI is here! Congrats, everyone!

Nvidia’s $12.9 billion Hugging Face deal could protect its open AI ecosystem, deepen ties with developers and keep the platform from rivals.

You can ignore it, tie it to unrelated company initiatives, or feign ignorance and simply not offer solutions that add to it. Or… you can adopt more above-board techniques such as Crawl/Walk/Run and the Reverse Farmer’s Rule, which reframe projects so ...

Inside OpenAI, coding agents are reshaping AI research. Explore early data on agent usage, experiment velocity, task complexity, and research acceleration.

Most of what an AI coding agent does for me isn't thinking. It's I/O.

Today, tech folk are scrambling to change their workflows to meet newly inflated 5X productivity quotas, while getting pummeled under the co...

The eleven lines of code that broke the internet, and what they teach us about our teams

Why shorter outputs can cost more, and how GitHub Copilot reduces wasted work across the complete coding task.

AI-generated tools are shifting shadow IT from SaaS sprawl to code sprawl. Here is how to secure your cloud infrastructure proactively.

In a report published Tuesday, Kaspersky said it first discovered NodeRabbit on a system in Afghanistan and later identified variants on systems in Egypt and Ethiopia.

Anthropic's Enterprise Frontier Safeguards are a response to customer feedback.

The memory industry is going through a supercycle for the history books. In 2023 the industry was at the bottom of its worst slump ever, facing negative industry-wide profits. Samsung cut production 50% and both SK Hynix and Micron posted negative gross margins! ChatGPT launched. Stargate alone would require

A multimodal request is not a text request with a picture attached. Walking the two side by side — payload, token accounting, cache identity, cost variance, and the encode stage — and showing what llm-d does about each difference. Sub-second TTFT p90 across a 5-40 req/s ladder and +44% throughput over round-robin on 16 H200s.

containerd 1.7, the final 1.x release line, reaches end of life on September 1, 2026 after three and a half years as an LTS. It is the container runtime under most Kubernetes nodes and Docker hosts — and the component almost no inventory lists. What changes, the 1.x to 2.x migration reality, and why 2.3 LTS (April 2028) is the landing zone.

A generation of technology workers grew up using it

What moving a Dockerized pipeline off my laptop taught me about containers, networking, and hidden assumptions.

I run a local LLM server on my M4 Pro Mac mini with 48 GB of RAM.

Apple Inc. renamed Lake Ontario “Lake America” on its maps service, following a similar move by Alphabet Inc.’s Google, in the wake of an order from US President Donald Trump.

Agents can finish the task without teaching you anything. Building expertise now has to be deliberate.

The Open-Source Elevenlabs alternative AI Voice Clone, Dub, Dictate, Transcribe, Audiobook creator and Voice workflow studio. - debpalash/VoiceStudio

Lawsuit: Anthropic’s torrenting totally screwed songwriters as AI songs top charts.

Memoryfields - a vastly simpler way to do agent memory

Pick a Mac for local LLMs by unified memory first: how much RAM and bandwidth you need, real tok/s by chip, the M5 Mac Studio timing, and exactly what to buy.

I’ve read and listened to pretty much everything I can get my hands on related to the Hugging Face attack. …

At the end of the previous article we watched a module fill in a struct net_device_ops (include/linux/netdevice.h:1436), hand it to register_netdev() (net/core/dev.c:11531), and walk away. That’s a promise: here are my function pointers, call me when you need the hardware. What we never got to see is the other half of the promise — who calls them, and what happens in between the wire and your program. So let’s go look. We’re going to follow one packet all the way in, and then one packet all the way back out.

Our Management Tip of the Day newsletter continues to be one of HBR’s most popular newsletters. In this article, we’ve compiled 10 of our favorite tips on learning from failure, from how to apologize to a customer to how to get your team back on track when a project is failing.