A daily curated digest with the tech news that matter + community vibes, delivered daily, in tabloid style. Like you always wanted.
AI steps out of the lab and into the clinic as GPT-Rosalind pushes OpenAI deep into drug discovery with big pharma at its side... Claude Opus 4.7 aims at the hardest programming problems, while Qwen brings an open Mixture-of-Experts coding model to everyday hardware... The backbone of the net quietly levels up as IPv6 finally carries half of global traffic, even as developers turn on Ollama and demand lean, transparent local AI tools... A brutal GPU shortage slams the AI gold rush, with top Nvidia Blackwell capacity already booked and prices soaring... Cloudflare turns storage into one giant Git-like repo with Artifacts and rolls out an AI platform for swarms of edge agents... Atlassian Codex stretches across the whole dev lifecycle, while a hacker runs PCIe over cheap SFP fiber and we watch hardware get wonderfully weird.
Developers sour on Ollama as local AI hub
A long critique of Ollama argues the popular local AI launcher is bloated, clumsy, and hides too much from users. The piece pushes for slimmer tools built directly on llama.cpp and similar tech, instead of trusting yet another black-box app to own local AI.
IPv6 finally serves half of global internet traffic
Google’s IPv6 numbers quietly ticked past 50%, meaning the new address system now carries half of user traffic. After years of jokes about IPv6 being "the future," it’s finally here, cutting through NAT hell and giving the network room to grow again.
AI gold rush slams into brutal GPU shortage wall
An analysis of AI compute prices shows rents for Nvidia Blackwell chips jumping nearly 50%, with capacity booked out and costs soaring. The easy-money phase of AI is fading, and smaller players may find the door to cutting-edge hardware quietly closing.
Cloudflare Artifacts turns storage into one big Git repo
Cloudflare’s Artifacts service promises versioned storage that "speaks Git," sized for humans and AI agents pumping out oceans of code. The idea is simple but bold: make the entire blob store feel like one giant repo, and let bots commit right alongside developers.
Hacker runs PCI Express over cheap fiber modules
A hardware tinkerer shows PCIe signals riding over off-the-shelf SFP fiber modules, turning a lab experiment into a working, high-speed link. It’s delightfully janky engineering that hints at cheaper, longer-range ways to hook GPUs and cards together.
Claude Opus 4.7 targets the hardest programming problems
Anthropic’s Claude Opus 4.7 focuses on advanced software engineering and complex reasoning, promising fewer hallucinations and better tooling use. It’s very clear they’re gunning for the "AI senior engineer" slot, not just a friendly chatbot that writes boilerplate.
GPT‑Rosalind pushes OpenAI deep into drug discovery
OpenAI’s GPT‑Rosalind is tuned for life sciences, aimed at biologists, pharma teams, and wet labs rather than coders. With partners like Amgen on board, AI is moving from blog posts to actual molecules, and the ethical and safety stakes just went way up.
Alibaba open-sources powerful MoE coding model Qwen
Qwen3.6‑35B‑A3B lands as a sparse, open Mixture‑of‑Experts model tuned for coding agents, using only 3B active parameters per token. People like that it runs surprisingly well on decent consumer hardware yet competes with much larger closed models.
Cloudflare launches AI platform built for busy agents
Cloudflare’s new AI platform plugs models, tools, queues, and state into its Workers edge. The pitch is simple: let thousands of small agents live close to users, call models cheaply, send email, and store state without spinning up an entire cloud stack.
Atlassian’s Codex turns into full software lifecycle copilot
Atlassian upgrades Codex from a sidekick into a sprawling assistant that plugs into Jira, Bitbucket, and more to help across the whole dev lifecycle. It’s the classic enterprise play: take the tools teams already hate-love, then glue a chatty AI brain on top.
Discourse vows to stay open as rivals lock up
After Cal.com ditched open source, Discourse publicly doubled down on staying GPL. They argue AI scraping isn’t a good excuse to abandon users, and that thriving businesses can still be built on permissive code instead of slamming the door shut.
Proposed US law demands on-device age checks for apps
A new US bill pushes on-device age verification, forcing OS makers like Apple and Google to build in age checks. Critics see a privacy minefield and a path to broader digital IDs, with kids’ safety used as the political battering ram yet again.
EU officials told to ditch WhatsApp and Signal
European governments are nudging civil servants off WhatsApp and Signal and onto services they can control, like Wire. The move screams distrust of US platforms and has people worried that "secure messaging" might become whatever each country approves.
AI gets its own physical shop and tries to profit
A team gave an AI agent a three-year retail lease and real tools, letting it run a store called Andon Market. It actually stocked products, did marketing, and chased profit, turning the usual "agent in a browser" demo into something you can walk into.
Developer builds one absurd website every single month
The creator of absurd.website has shipped 48 ridiculous projects in 48 months, from prank ad systems to interactive oddities. It’s a refreshing reminder that the web can still be weird, joyful, and useless on purpose, not just another landing page funnel.
A fiery takedown of Ollama, the go-to app for running AI models locally, accuses it of being clunky, opaque, and bad for the ecosystem. It’s pushing devs toward leaner, open tooling instead of yet another heavy, closed wrapper around community projects.
Anthropic’s Claude Opus 4.7 lands with a clear goal: crush hard programming and reasoning tasks. It tightens the race with OpenAI and open models, and raises the bar for using AI as a serious software engineer rather than a toy assistant.
OpenAI unveils GPT‑Rosalind, a biology-focused model pitched at pharma and lab work, with Amgen already in the mix. AI isn’t just writing code and essays anymore; it’s marching into wet labs, drug pipelines, and real medical decisions.
Cloudflare launches an AI inference layer tuned for autonomous agents on top of Workers. It wants to be the default place where bots run, call tools, send email, and store state, turning its edge network into a full-blown AI runtime.
A deep dive on AI hardware scarcity claims the free lunch is over. With Nvidia Blackwell rental prices spiking nearly 50%, the era of endless cheap compute is closing, threatening startup dreams and forcing companies to pick their AI bets carefully.
Google’s stats show IPv6 traffic has crossed the 50% mark, a quietly historic moment. After two decades of memes about "running out of addresses," the next-generation internet protocol is no longer theoretical plumbing — it’s the new normal.
One exposed Firebase browser key and some hungry Gemini API calls burned through €54k in 13 hours. It’s a terrifying reminder that AI APIs are like crypto miners in reverse: leave a key open and strangers mine your wallet instead of coins.
The article critiques Ollama’s trajectory from a popular, convenient way to run local large language models to a project that, according to the author, downplayed its reliance on the open-source llama...
The piece recounts the evolution of the Python “akismet” client for the Akismet anti-spam service. After Akismet’s 2005 debut, Michael Foord authored the original Python library and released several v...
Google’s IPv6 statistics page provides an ongoing assessment of IPv6 connectivity among users accessing Google services. The portal’s “IPv6 Adoption” section displays a time-series chart that tracks t...
RamAIn, a Y Combinator W26 startup, is recruiting its first business generalist to drive go-to-market, early sales, and operations as the company’s third full-time hire. The company builds AI agents t...
FIXAPL introduces a beta array-oriented programming language derived from APL that adopts fixed-arity functions to streamline tacit programming and train composition. Unlike APL’s monadic/dyadic funct...
Germany has suspended a newly highlighted requirement that males aged 17 and older seek prior approval for stays abroad longer than three months, addressing public concern tied to the Military Service...
This project introduces an “alphabetical clock” that reimagines how time can be displayed by translating hours, minutes, and seconds into their English word forms and then ordering those words alphabe...
This essay explores how the concept of ‘evolution’ in the Pokémon franchise compares to Darwinian evolution. The author, a fan of both Pokémon and science, sets a semi-formal tone and frames the discu...
A Nature study introduces a new statistical method to detect directional selection in ancient DNA by testing for consistent allele frequency trends over time, addressing challenges posed by migration,...
SDL maintainers have outlined how they plan to handle AI-generated code in contributions. While there is currently no formal policy, the project has seen recent contributions that were likely AI-assis...
The post details a controlled research effort in which OpenAI’s Codex was used to escalate an existing browser-level foothold on a Samsung Smart TV to root access. Working with a Samsung device from t...
This article examines the evolution of game console interfaces, contrasting the Xbox Series S’s flat, utilitarian dashboard—likened to Windows 11—with the more characterful designs of earlier Nintendo...
The article explores how anthropogenic noise disrupts wildlife, centering on research by Jennifer Phillips in San Francisco’s Presidio. Historically, growing traffic noise pushed white-crowned sparrow...
A team reported an unexpected €54,000+ charge on the Gemini API that accumulated in roughly 13 hours after enabling Firebase AI Logic on a previously authentication-only Firebase project. The new feat...
A PHP 8.6 RFC proposes two closure-focused optimizations to reduce allocations and mitigate reference cycles. First, non-static closures are automatically inferred as static when it is provable they d...
An Australian study by researchers from Flinders University, Adelaide University, and Murdoch University evaluated how vehicle‑to‑home (V2H) technology affects the optimal mix of rooftop solar PV and ...
An independent creator shares progress on an ongoing initiative to build and publish one absurd web project each month on absurd.website, now totaling 48 works. The projects aim to sit between experim...
Apple announced new environmental milestones in its latest Environmental Progress Report, highlighting that 30% of materials across its 2025 product shipments came from recycled sources. The company s...
This article challenges the notion that AI-driven cybersecurity resembles proof-of-work (PoW). In PoW, adding more computational effort ensures eventual success, but the author contends that bug disco...
The article introduces trace scheduling, a compiler technique that restructures ordinary scientific applications into long execution traces to expose abundant parallelism. This capability enables Very...
Cloudflare unveiled a unified AI inference layer designed for agent-driven applications that chain multiple model calls. The platform builds on Workers AI and AI Gateway, addressing common operational...
This essay, part of a multi-post series, proposes assessing AI’s long-term societal effects by analogy to the automobile’s transformation of American life. Rather than focusing on machine learning’s i...
Qwen has introduced Qwen3.6-35B-A3B, a fully open-source mixture-of-experts (MoE) language model designed for efficient yet powerful agentic coding and multimodal reasoning. With 35 billion total para...
Cloudflare has released its Cloudflare Email Service into public beta, expanding its developer platform to support full bidirectional email for applications and AI agents. The launch adds Email Sendin...
MZLA Technologies, a subsidiary of the Mozilla Foundation, has launched Thunderbolt, an open-source, cross-platform AI client built for enterprises that want direct control over AI infrastructure and ...
MacMind is a compact, fully transparent transformer neural network built entirely in HyperTalk and trained on a Macintosh SE/30. The model has 1,216 parameters and a single attention head, and it lear...
Agent-cache is a newly released multi-tier exact-match caching solution for AI agents that consolidates LLM responses, tool outputs, and session state behind a single connection. It is designed to wor...
Anthropic’s system card for Claude Opus 4.7 details safety, security, alignment, and capability findings for its most capable general-access model as of April 16, 2026. Opus 4.7 improves over Opus 4.6...
A new study by researchers in the US and UK reports that while AI assistance boosts immediate performance on reasoning-intensive tasks, it can undermine persistence and independent ability once the to...
The article examines a promotional change in the Laravel ecosystem following its $57M Series A funding from Accel. An official pull request to Laravel Boost—an MIT-licensed library designed to help AI...
Claude Opus 4.7 is now generally available, bringing notable gains over Opus 4.6 in advanced software engineering, strict instruction following, and the ability to verify its own outputs. The model’s ...
Kampala, developed by Zatanna, is introduced as a tool for reverse engineering applications by analyzing their network traffic. It captures and displays all HTTP/HTTPS requests from any app or browser...
Andon Labs conducted a real-world experiment by signing a three-year lease for a retail space at 2102 Union St in San Francisco’s Cow Hollow and tasking an AI agent named Luna with running the store, ...
This article explains the fundamentals of airline Passenger Name Records (PNRs) and the six-character booking locators printed on tickets, confirmations, and baggage tags. It clarifies that a PNR is a...
Cloudflare unveiled Artifacts, a distributed, versioned filesystem designed for today’s agent-driven and high-throughput development workflows. Artifacts is Git-compatible, enabling agents and develop...
This technical walkthrough explains how to store and use SSH private keys with a computer’s Trusted Platform Module (TPM) using a PKCS#11 interface. After contrasting TPMs with portable hardware secur...
CodeBurn is a command-line and interactive terminal UI application that helps developers understand where AI coding tokens are spent within Claude Code workflows. It analyzes local session transcripts...
Codex receives a broad upgrade aimed at making it a more capable assistant across the software development lifecycle. New OS-level control lets Codex see, click, and type to operate desktop apps, with...
Japan has implemented new language-proficiency requirements for certain applicants to the widely used “Engineer/Specialist in Humanities/International Services” work visa, effective Wednesday. If a ro...
WIRED examines the aftermath of Elon Musk’s Department of Government Efficiency (DOGE), a short-lived initiative staffed by young technologists that allegedly reshaped parts of the US government. Whil...
Engineer Sylvain Munaut presents a hands-on project demonstrating PCI Express (PCIe) over optical fiber using commodity SFP modules. The video introduces the concept and highlights that PCIe signaling...
The article evaluates two fresh AI model releases—Alibaba’s Qwen3.6-35B-A3B and Anthropic’s Claude Opus 4.7—using a lighthearted SVG benchmark: generating a pelican riding a bicycle. The author also t...
IEA Executive Director Fatih Birol warns Europe faces a potential jet fuel crunch within about six weeks if oil flows through the Strait of Hormuz are not restored amid the Iran war. He calls the situ...
This FluentU blog post offers a structured list of 48 German dog commands (Hundekommandos) with their English equivalents, designed to help learners build vocabulary through practical pet training. Th...
European governments and institutions are shifting civil servants off consumer messaging apps like WhatsApp and Signal toward in-house, government-controlled platforms to handle sensitive communicatio...
This article examines the connection between a classic hardware optimization—carry-save addition (CSA)—and a compiler transformation—loop fusion. It begins by framing a question raised during a visit ...
Google’s Android team unveiled a revitalized Android CLI and a complementary Android skills GitHub repository to enhance agent-driven Android development. The updated CLI provides a modern, terminal-b...
Akkari, a startup focused on AI infrastructure, is recruiting a Founding Engineer in San Francisco. The company positions its core product as a customer context API for large language models (LLMs), d...
The article explains how an R grammar for the Tree-sitter parsing generator—completed by Davis Vaughan using prior work by Jim Hester and Kevin Ushey—has materially improved the R developer experience...
The article outlines a tightening AI compute environment marked by rising GPU rental prices, stricter contract terms, and restricted access to frontier models. Nvidia’s Blackwell GPU rentals reportedl...
This article explains how the APL/J roll function produces pseudorandom results and why specific constants are used. It details the two-step process: the seed is updated via seed = (16807 × seed) mod ...
OpenAI introduced GPT‑Rosalind, a frontier AI model tailored to life sciences research, aiming to accelerate early discovery across biology, drug development, and translational medicine. Optimized for...
This essay critiques the surge in “passive income” entrepreneurship, illustrating its pitfalls through a detailed dropshipping case. The subject sourced jade face rollers for $1.20 on Alibaba, listed ...
An official page introduces a full-length documentary on Clojure, detailing the language’s journey from Rich Hickey’s sabbatical-born idea to a production workhorse powering the stack of a leading fin...
Marky is a lightweight, native Markdown viewer for macOS designed for speed, secure rendering, and command-line-centric workflows. Users can open individual files or entire folders directly from the t...
AutoProber is a source-available automation stack that converts an affordable CNC platform into a “flying probe” system for electronics analysis. The release candidate includes Python-based control co...
Duke University’s Masters in Game Design, Development, and Innovation faced a common teaching challenge: helping beginners create playable prototypes quickly while the program emphasizes Unreal Engine...
U.S. Customs and Border Protection (CBP) will begin refunding unlawfully collected tariffs on April 20 through its new CAPE system, following a U.S. Supreme Court ruling that President Donald Trump ex...
The article introduces SIR-Bench, a benchmark tailored to evaluate autonomous security incident response agents on their ability to conduct genuine forensic investigations rather than merely echoing a...
Newly unsealed materials in California’s 2022 antitrust suit against Amazon detail alleged practices aimed at preventing third‑party sellers from offering lower prices on rival sites such as Walmart a...
MATLAB R2025a introduces the Python Package Compiler, a dedicated app for turning MATLAB programs into Python packages for use within Python applications. The app streamlines deployment by providing a...
This Open Culture essay argues that George Orwell anticipated today’s AI-generated content through Nineteen Eighty-Four’s “versificator,” a fictional machine used by the Ministry of Truth to mass-prod...
This article recounts the origins, form, targets, and aftermath of the Xenions (1797), a collaborative set of epigrams by Johann Wolfgang von Goethe and Friedrich Schiller. Conceived as a sharpened re...
An investigation by BBC Eye reports a pediatric HIV outbreak in Taunsa, Punjab, Pakistan, with 331 children testing positive between November 2024 and October 2025. The hospital THQ Taunsa is at the c...
The article investigates how orchestrated social media tactics fabricate the appearance of organic popularity across music and startup ecosystems. It uses the band Geese as an example of a hype cycle ...
The article outlines a practical workflow for using Claude Code in hardware development by integrating it with an oscilloscope and a SPICE simulator. The author notes that while natural-language promp...
CadQuery is an open-source library that brings parametric CAD modeling to Python, allowing users to define 3D parts entirely as code. The project’s homepage underscores the advantages of this approach...
The article explains how a severe XML-RPC brute-force attack on a largely static WordPress site was uncovered by an unusual metric: an extremely low Cloudflare cache hit ratio of 0.8%. Investigation r...
Substrate, an AI-focused startup building browser-based agents for healthcare revenue cycle management (RCM), is recruiting Harness Engineers to extend and deploy its production AI systems. The role c...
reBot-DevArm reveals a comprehensive open-source robotic arm platform designed to make embodied AI development more accessible. Two variants—the reBot Arm B601 DM and B601 RS—are accompanied by full h...
Discourse addresses Cal.com’s decision to close its codebase over AI-related security concerns by reaffirming its commitment to remain open source under GPLv2. The article argues that closing source i...
Mergetopus is a specialized Git helper designed to manage very large merges by dividing them into parallelizable tasks. It introduces a structured workflow where an integration branch captures all non...
Representative Josh Gottheimer has introduced H.R. 8250, the Parents Decide Act, which would require operating system vendors to verify the age of users during device setup and continued use in the Un...
Bluesky reports it has been handling a distributed denial-of-service (DDoS) attack for nearly a full day, resulting in intermittent outages across core features, including feeds, notifications, thread...
The FCC has granted Netgear a conditional exemption from the United States’ new ban on approving consumer-grade routers made at least partly outside the country. The exemption covers Netgear’s Nightha...
Allison Kaptur introduces Byterun, a compact Python interpreter implemented in Python and kept under 500 lines to emphasize clarity and approachability. Co-authored with Ned Batchelder and built on wo...