news

OpenAI safety leader quits amid major AI industry shifts

04 Oct 2026
We're going to need default hard budget caps on pretty much everything

A recent tech commentary highlights the growing necessity for digital services to implement default hard budget caps. The piece argues that as technology costs rise in the coming months, making these financial limits a standard product feature will be essential to prevent unexpected user expenses.

Comments
Bob Cringely Has Died

Prominent tech journalist and author Bob Cringely, best known for his influential Infoworld columns, the book Accidental Empires, and his PBS documentary Plane Crazy, has passed away. The tech community is mourning his death while celebrating his foundational impact on the industry and his pioneering journalism.

Comments
Treachery in the Rodin Museum 3D scan verdict
Treachery in the Rodin Museum 3D scan verdict

A recent court verdict involving the Rodin Museum has sparked controversy over the intellectual property rights of 3D scanned public domain artifacts. The ruling highlights a confusing intersection of digital replication and copyright law, demonstrating how strict legal interpretations can sometimes undermine the intended spirit of the law.

Comments
Hole Punch: Sling your spaceship around gravitational fields

Hole Punch is a physics-based space game where players navigate their spaceship home by creating black holes. Users must strategically manipulate gravitational fields to slingshot their vessel safely through the cosmos.

Comments
The work by Valve's Timur Kristóf on improving old AMD GPUs on Linux
The work by Valve's Timur Kristóf on improving old AMD GPUs on Linux

Over the past year, Valve developer Timur Kristóf has implemented multiple improvements to the Linux AMDGPU kernel driver to better support decade-old GCN 1.0 and 1.1 graphics cards. These updates are specifically designed to enhance gaming performance and general task handling for users running older AMD hardware on Linux.

Comments
Inside Anthropic's Quest to Instill Morality into Its A.I. Models

Anthropic is intensifying its efforts to embed ethical guidelines and moral frameworks directly into its artificial intelligence models like Claude to ensure they align with human values. This initiative highlights the broader tech industry challenge of defining, measuring, and programming objective morality into complex machine learning systems to prevent harmful outputs.

Comments
Reasons I didn't become an EMT, ranked
Reasons I didn't become an EMT, ranked

In a humorous personal essay, the author details 21 reasons for delaying their EMT licensing. These excuses are ranked and sorted loosely by their level of ridiculousness, offering a lighthearted look at career procrastination.

Comments
Celebrating the 100th birthday of the kidney donated to him as a teenager
Celebrating the 100th birthday of the kidney donated to him as a teenager

A Webster man is celebrating the 100th birthday of the kidney his mother donated to him during a 1978 transplant at Strong Hospital. Ray Vetuskey was just 16 years old when he received the life-saving organ. This remarkable medical milestone marks a full century since the kidney was born.

Comments
So You Think You Could Be an Electrician?
So You Think You Could Be an Electrician?

This article examines the irony of white-collar professionals urging younger generations to pursue careers in the skilled trades. It highlights the stark disconnect between privileged advice-givers and the physical realities of blue-collar jobs such as electrical work.

Comments
Kolibri: A Sovereign Open-Weight Model
Kolibri: A Sovereign Open-Weight Model

The newly released Kolibri is an English-German mixture of experts model optimized for sovereign and mission-critical workloads. It features 78 billion total parameters with 3 billion active, a one million token context window, and fully open weights under the Apache 2.0 license.

Comments
OpenAI safety leader quits, warning AI company's culture is 'broken'
OpenAI safety leader quits, warning AI company's culture is 'broken'

OpenAI safety leader David Robinson has resigned from the company, warning that its internal culture is broken. His departure adds to a growing chorus of industry insiders urging the technology sector to exercise greater caution and prioritize safety in the rapid development of artificial intelligence.

Comments
Math's pedagogical curse – Grant Sanderson [video]
Math's pedagogical curse – Grant Sanderson [video]

In a recent presentation at the 2023 Joint Mathematics Meetings, popular math educator Grant Sanderson addressed the pedagogical curse that hinders effective mathematical communication. He outlined actionable steps the mathematics community can take to overcome these conceptual barriers and improve how complex ideas are taught to broader audiences.

Comments
Getting the most out of Opus 5.5 in Claude and Claude Code
Getting the most out of Opus 5.5 in Claude and Claude Code

A new guide details how to maximize the capabilities of the Opus 5.5 model within both the standard Claude interface and Claude Code. The resource outlines essential techniques for effective prompting, steering extended execution runs, and accurately verifying final outputs.

Comments
Agents don't need memory, they need documentation

A new perspective on AI agent design argues that structured project documentation is far more effective than conversational memory for maintaining context. By prioritizing comprehensive documentation over retrieval from past interactions, developers can ensure more accurate and reliable autonomous task execution.

Comments
Show HN: Thoreau BASIC – What if BASIC hadn't gone out of fashion?

Thoreau BASIC 3.2 is a free 64-bit interpreter that modernizes classic GW-BASIC syntax for Windows and bootable UEFI environments with a native x64 JIT compiler and multicore processing. The software aims to revive the nostalgic programming language by adding contemporary features like 24-bit graphics, networking, and extensive multimedia playback capabilities.

Comments
FTL: A new operating system for clouds
FTL: A new operating system for clouds

A new operating system named FTL has been introduced, specifically designed to manage and optimize cloud computing environments. This innovative platform aims to streamline cloud infrastructure and improve overall system performance. By providing a dedicated OS for the cloud, FTL seeks to enhance efficiency and scalability for modern digital applications.

Comments
Your body of work thinks back at you

A recent essay explores how reviewing a personal photographic archive can reveal hidden patterns and underlying themes within one's own body of work. The author suggests that our own curiosity drives this reflection, allowing unnoticed details to culminate in profound personal realizations. However, skeptics counter that perceiving these consistent threads may simply be a result of apophenia and confirmation bias.

Comments
New York City should carefully measure a new tree
New York City should carefully measure a new tree

New York City's tallest recorded tree is currently the Queens Giant. This tree was measured in 2004 at 133.8 feet tall with a 19-foot circumference. The city is now being urged to carefully measure a new tree to see if this long-standing record has been broken.

Comments
Surely you have ultra-wideband radios on your bins too?

A maker recently detailed a DIY project where they equipped their household trash bins with ultra-wideband radios for precise tracking and automation. Although the creator admits to massively over-engineering a trivial problem, the build serves as an impressive showcase of advanced UWB technology in everyday smart home setups.

Comments
Show HN: Pi pod – Run your pi coding agent in sandboxes on your own server

A new tool called Pi pod enables developers to run their Pi coding agent in secure, isolated sandboxes directly on their own servers. The platform uses cloud pods to provide a self-hosted environment for safely executing and managing coding sessions.

Comments
Three AI agents, two countries, and one uneven world wide web
Three AI agents, two countries, and one uneven world wide web

A new evaluation compares GPT, Claude, and Muse across two countries to assess their capabilities in multilingual research, source access, human-in-the-loop workflows, and observability. The study reveals significant disparities in how these AI agents navigate the uneven global web and process diverse linguistic contexts. Ultimately, the findings highlight the ongoing challenges developers face in achieving truly equitable AI performance across different geographic regions.

Comments
Federal judge calls Flock 'indiscriminate mass surveillance'
Federal judge calls Flock 'indiscriminate mass surveillance'

A federal judge ruled that a sheriff's deputy violated a woman's Fourth Amendment rights by using Flock's license plate tracking technology to locate her vehicle without a warrant. Condemning the automated system as indiscriminate mass surveillance, the decision marks a significant legal pushback against law enforcement's use of such tracking tools.

Comments
Woking Electrical Control Room (2016)
Woking Electrical Control Room (2016)

A 2016 feature highlights an abandoned railway electrical control room located in Woking, Surrey. Once a critical hub for regional transit infrastructure, the now-unused facility offers a rare glimpse into historical electrical control systems left behind by modern upgrades.

Comments
We want you to build the next Git platform on Cloudflare
We want you to build the next Git platform on Cloudflare

Cloudflare has launched a developer competition to build a next-generation Git platform designed for the era of AI agents. The initiative encourages the use of Cloudflare's newly open-beta Artifacts, which includes Workers bindings, data jurisdiction controls, and event subscriptions for repository changes.

Comments
Docker has always used microVMs (well since 2016)
Docker has always used microVMs (well since 2016)

A recent article clarifies that Docker Desktop has been running containers inside minimal, purpose-built microVMs since 2016. The piece explores the technical reasoning behind this long-standing architectural decision and outlines the key benefits it has delivered to the platform.

Comments
Apple's "Clean Design" Is Stupidity When It Comes to Hiding Fire Extinguishers
Apple's "Clean Design" Is Stupidity When It Comes to Hiding Fire Extinguishers

Apple's commitment to minimalist design has led to the concealment of fire extinguishers in its facilities to maintain a sleek aesthetic. However, this approach violates OSHA and NFPA regulations that mandate clear identification and easy access to emergency equipment. Ultimately, the controversy highlights the dangers of prioritizing polished design over basic safety compliance.

Comments
RetailReady (YC W24) Is Hiring
RetailReady (YC W24) Is Hiring

RetailReady, a Y Combinator-backed AI supply chain compliance startup, is hiring for its in-person San Francisco team after raising $6.6 million and quadrupling its revenue. The company is seeking candidates to lead software deployments and collaborate with engineering to modernize antiquated retail supply chain systems for brands and warehouses.

Comments
RSS Feed Best Practices (2022)

A community discussion surrounding RSS feed best practices clarifies the limited support for complex styling across readers, recommending publishers rely on inline CSS, semantic HTML, and whitelisted iframes while strictly avoiding JavaScript. Participants also emphasize the importance of maintaining reliable feeds, adopting the Atom 1.0 standard, and using tag URIs for entry IDs to ensure maximum compatibility. Ultimately, the consensus encourages developers to keep their feeds intact and functional despite major platforms frequently violating these technical standards.

Comments
Memory-Safe WebP Decoding
Memory-Safe WebP Decoding

A new memory-safe approach to WebP decoding has been introduced to significantly enhance both performance and security. This update ensures faster and safer image processing for all users across different platforms. By prioritizing memory safety, the improvement also helps mitigate potential vulnerabilities in web rendering.

Comments
Former SR-71 engineer talks NASA's Blackbird revival program
Former SR-71 engineer talks NASA's Blackbird revival program

A former SR-71 engineer recently shared insights into NASA's initiative to restore the iconic Blackbird aircraft #844 to flightworthy status. The update detailed the extensive technical requirements needed for the revival and explored potential future high-speed research missions for the agency.

Comments