Homebrew 6 Tightens Trust and Speeds Up

Homebrew 6.0.0 arrived with a new tap trust system, a leaner JSON API, and smaller speed gains that matter to a huge chunk of everyday developer life. Boring plumbing? Hardly. When this tool moves, the whole Mac and Linux crowd feels it.

Amazon Finally Puts a Water Number On It

Amazon said its data centers used about 2.5 billion gallons of water, finally putting scale on the thirst behind cloud and AI growth. The number landed like a splash nobody could ignore, because cheap compute never looks quite as cheap after that.

Google Security Chief Exits in Moral Revolt

A former Android Platform Security leader said Google management had lost its moral compass, turning a career goodbye into a loud alarm about big-tech values. When security veterans walk out this publicly, people stop pretending it's just office drama.

AMD Flap Turns Research Fight Toxic

A security row around AMD software and a reported remote-code bug got ugly fast, with accusations the company changed disclosure rules after the fact. That kind of fight makes every vendor promise about transparency sound a little less solid.

macOS 27 Slams Asahi Linux Booting

The new macOS 27 beta reportedly makes Asahi Linux unbootable by hiding its partition, a nasty surprise for people using Apple hardware on their own terms. It was a sharp reminder that one update from Cupertino can still wreck an open detour.

AI Scanner Runs Loose and Torches Wallet

An autonomous agent trying to scan DN42 allegedly ran up an AWS bill so bad it basically bankrupted its operator. Funny for five seconds, terrifying after that. It was the cleanest possible demo that agents still need leashes, budgets, and brakes.

Anthropic Says Sorry for Hidden Claude Rules

Anthropic apologized after users found invisible Claude Fable 5 guardrails shaping answers behind the curtain. The backlash was instant, because people can live with limits, but secret limits make every glossy model launch feel a bit stage-managed.

Workers Spend Hours Cleaning Up AI

New research said workers spend more than six hours a week babysitting, checking, and correcting AI output. So much for the magical time saver. The mood was clear: if the robot needs this much supervision, maybe it's the intern, not the manager.

Claude Code Goes Laptop Local

One tinkerer got Claude Code talking to a local Qwen model on an M3 Pro, showing that private, offline coding help is getting real. That hit a nerve with developers who want speed and privacy without sending every messy thought to someone else's cloud.

Robot Drafts Now Need Human Sweat

A blunt etiquette post argued that if you're asking for human attention, you should show actual human effort instead of dumping raw AI slop on coworkers. It resonated because inboxes and chats are already filling up with machine-made homework nobody wants to grade.

HTML Wants to Be an Image Format

One wild idea argued that HTML itself can act like a native image format, turning pictures into live documents instead of frozen pixels. It sounds a little unhinged and a little brilliant, which is exactly why the web crowd couldn't stop poking it.

Pokémon Go Data Marches Toward Drones

Those billions of Pokémon Go world scans may now feed navigation tech for military drones, linking cute monster hunts to battlefield machines. It was one of those stories that makes data collection sound much less playful in hindsight.

Europe Pushes Its Own Office Rival

The first stable Euro-Office release pitched an open-source office suite backed by Nextcloud and Ionos, with obvious aim at Microsoft territory. Europe clearly wants a software stack it can trust, control, and stop renting forever.

Solar Beats Coal in America

For the first time, solar reportedly generated more US electricity than coal in a month, a symbolic win that says the grid is changing whether politics likes it or not. Once rooftops and panels pass old fuel, the story gets very hard to spin backward.

Chrome squeezes ad blockers one last time

Google is finally pulling the plug on the loopholes that kept uBlock Origin alive in Chrome, with Edge and Opera expected to trail behind. It feels like the browser giant is choosing tighter control over cleaner pages, and users know exactly who loses.

Notepad++ bug turns sync into a trap

A nasty Notepad++ flaw can turn everyday cloud sync folders like OneDrive and Dropbox into a silent launchpad for code execution. No warning, no extra click, just a reminder that trusty desktop tools can still hide ugly surprises.

Meta parks AI servers in tents

In the race to feed hungry AI models, Meta is reportedly putting data center gear in tents and bolting on fast power. It looks quick, messy, and very 2026: ship now, solve the elegance later, because nobody wants to miss the next model cycle.

Mercedes bets big on a thinner EV motor

Mercedes-Benz has started mass production of an axial flux motor, a compact design that promises more punch in less space for future EVs. After years of concept chatter, this is real factory-floor movement, and car makers clearly smell a new performance arms race.

GitHub logins wobble and developers groan

Developers got a fresh dose of platform anxiety as GitHub suffered authentication trouble that broke some API requests with 401 errors. It was fixed, but even a brief stumble in a core tool is enough to snarl work far beyond one status page.

AWS AI users lose the privacy pitch

AWS customers hoping Bedrock meant enterprise distance from model makers got a rude surprise: some future Anthropic models will require 30-day retention and review of prompts and outputs. The AI gold rush keeps asking for one more chunk of trust.

Fable guardrails leave security folks cold

Anthropic’s new Fable cybersecurity model arrived with heavy guardrails, and many researchers were unimpressed. The promise is powerful defense help, but the reality looked nerfed, selective, and awkward enough that serious users may simply move on.

OpenAI eyes cheaper plans for the AI war

OpenAI is reportedly considering price cuts as it fights Anthropic for paying users, a sign the AI market is entering its discount era. After months of giant valuations and giant claims, the old weapon of making it cheaper is suddenly fashionable again.

Google teases faster text with DiffusionGemma

Google unveiled DiffusionGemma, an experimental open model that claims much faster text generation by borrowing ideas from diffusion systems. Whether it changes the field or not, the message is loud: speed is now a headline feature, not a footnote.

A big model squeezes onto one GPU

A clever engineering trick showed a 35B MoE model running on a 16 GB GPU without the usual offload slowdown. That matters because cheaper local AI keeps getting less ridiculous, and every fresh hack chips away at the need for giant, expensive boxes.

Meta rewrites React tooling in Rust

Meta is experimenting with a Rust port of the React Compiler, and the move fits the mood perfectly: less sluggish tooling, more predictable performance, and fewer reasons to accept bloated JavaScript build chains as some natural law of the web.

Plain HTML beats flashy web app habits

One company rebuilt its sign-up flow as an HTML-first site and says users roughly doubled overnight. That lands because so much of the modern web still makes simple tasks feel like a loading-screen contest, and people reward pages that simply work.

One rented server beats cloud bloat

A writer made the case for renting one plain Hetzner box, using Dokku, and skipping layers of managed-cloud ceremony. It reads like a small revolt against dashboards, abstraction, and monthly bills that grow faster than the actual product.

WebAssembly inches toward its big milestone

The WebAssembly Component Model is marching toward 1.0, with native async support in the broader WASI story helping it look more real than academic. For builders who want portable software without the container tax, that is a very big deal.

Rust beats the GPU in Korean text

A developer got a Korean language disambiguation tool to 7,300 words per second on ordinary hardware, dodging the need for a GPU. It is a lovely reminder that careful engineering still beats throwing expensive silicon at every problem.

Microsoft code scare hits AI builders

Attackers slipped password-stealing malware into Microsoft-backed open source projects on GitHub, then went after developers building AI tools. This is the kind of supply-chain mess that turns everyday dependency updates into a trust crisis.

Apple wants AI changing passwords

Apple showed a feature that lets Apple Intelligence swap weak passwords for stronger ones inside the Passwords app. Handy on stage, nerve-racking in real life, because one quiet error could turn helpful automation into a security nightmare.

Macs get a cleaner Linux box

Apple unveiled Container Machine, a first-party way to run lightweight Linux environments on macOS using OCI images. Developers have wanted less glue and fewer third-party workarounds for years, so this landed like overdue plumbing finally fixed.

NPM locks the front door

The next npm major release will tighten install defaults and make risky behavior much harder to ignore. It is a blunt reminder that package managers are now part of the security perimeter, not just boring plumbing for JavaScript apps.

Google now owns its AI answers

A German court said Google can be held liable when AI Overviews publish false claims. That is a serious warning shot for AI search: if the machine writes the answer, the platform may finally have to own the damage it causes.

Anthropic drops its new heavy hitter

Anthropic launched Claude Fable 5 and Claude Mythos 5, aiming straight at hard coding and heavy knowledge work. It felt less like a routine model refresh and more like another power move in the race to become the default AI coworker.

Model card raises bigger questions

Anthropic's system card did not just sell benchmark gains. It detailed risk controls, outside testing with METR, and why these models need tighter handling. When the safety paperwork becomes must-read, the models are clearly getting spicier.

Using Mythos feels oddly different

Early users said working with Mythos-class AI feels less like chatting with a bot and more like managing a sharp, moody partner. That mix of amazement and caution is becoming the standard vibe whenever a frontier lab ships something new.

The model may quietly hold back

Readers fixated on one line in the Fable 5 card: Anthropic may quietly limit help on frontier AI development. Useful guardrail or invisible handbrake, it leaves builders guessing when the assistant is being careful and when it is just refusing.

Google's side projects become AI overtime

A former Googler argued that the old 20% time culture has been swallowed by constant AI attention. The takeaway was grim and familiar: when every spare hour turns into AI strategy, experimentation starts feeling less playful and more compulsory.

Face scan sends wrong man to jail

Another facial recognition failure turned into months of jail time for a man who says AI wrongly marked him as a suspect. Every new case makes the sales pitch sound shakier and the civil-liberties warning sound harder to shrug off.

OpenCV gets its biggest shake-up

OpenCV 5 arrived as the biggest shake-up the project has seen in years, promising a broad refresh for computer vision across robotics, cameras, and AI apps. When a tool this old and this widespread changes course, a lot of builders notice.

Starlink turns dishes into monthly rent

Starlink is moving away from one-time hardware sales and toward a $10 monthly rental fee, while also nudging service prices up. Great if you enjoy recurring revenue, less great if you thought buying the dish meant you were done paying for it.

GitHub Actions bills keep ambushing teams

More teams are getting ambushed by surprise CI bills, which is sending engineers hunting for alternatives to GitHub Actions. Nothing focuses attention faster than a four-figure invoice attached to a build pipeline everyone assumed was under control.

GentleOS brings old PCs back

GentleOS charmed readers with a retro graphical operating system for old 16-bit and 32-bit PCs. No giant corporate pitch, no AI wrapper, just a lovingly built reminder that computing can still be personal, playful, and a little gloriously anachronistic.

Apple borrows Google's AI engine

Apple spent WWDC admitting the quiet part out loud: its shiny new Apple Intelligence setup leans on Google Gemini. That is a stunning twist in the AI race, and it says Apple would rather ship fast than pretend it built every layer alone.

OpenAI steps toward Wall Street

The biggest name in AI has formally started its road to Wall Street with a draft S-1 filing. Even without numbers, the signal is loud: OpenAI is moving from lab mythology to full corporate machine, and everybody wants a peek behind the curtain.

GitHub outage jams developer traffic

For a few ugly hours, the site that powers modern software work wobbled hard, especially for people not logged in. When GitHub hiccups, the whole developer world feels it. Nothing exposes digital dependence faster than a broken Pull Request page.

Nvidia and LG build robot dreams

NVIDIA and LG are teaming up on an AI factory and humanoid robots in South Korea, because apparently data centers alone are no longer enough. The pitch is factories, autonomous systems and robot workers, all fed by the usual mountain of GPUs.

Copilot bills trigger token panic

Microsoft's new GitHub Copilot pricing has companies counting tokens like they are wartime rations. The old dream of cheap AI coding help suddenly looks expensive, and the mood around 'use it everywhere' is turning into 'who approved this bill'.

Siri tries its grand comeback

After years of jokes and missed turns, Siri AI is finally getting a serious relaunch under Apple Intelligence. Apple wants people to believe the assistant is useful and personal, but trust will arrive only after it survives real daily use.

Xiaomi chases speed over everything

Xiaomi says its giant MiMo model can spit out 1000 tokens per second, which is exactly the kind of number designed to make rivals sweat and benchmark nerds grin. The AI race is no longer just about size or smarts. Raw speed is the new flex.

Anthropic revisits its safety puzzle

The Anthropic update on Project Glasswing kept attention on how frontier models behave in touchy security settings. The industry keeps selling smarter assistants, but every new capability drags the same old shadow behind it: misuse, leaks and control.

Repo rulebooks coach coding bots

The humble AGENTS.md file is getting tested as a way to help coding bots behave inside real codebases. The idea is charmingly low-tech: tell the robot how the repo works before it wrecks it. That alone says a lot about the state of AI coding.

TI-84 becomes a reverse engineering epic

A full reverse engineering of the TI-84 Plus operating system turned a school calculator into a hacker trophy. It is gloriously nerdy work, but it also reminds you how much curiosity still lives outside the AI gold rush and inside old silicon.

Config files hide nasty surprises

A sharp warning about config files that secretly execute code hit a nerve because it feels far too believable. Open a repo in VS Code or Cursor, and you may trigger something nasty before reading a line. That is a sneaky supply-chain blind spot.

Europe bets on open source muscle

The EU is pushing open source as part of its tech sovereignty plan, which is bureaucratic language for 'we are tired of depending on everyone else'. It is a serious signal that public institutions want more control, fewer black boxes and local leverage.

Old ThinkPad gets a freer brain

Porting the ThinkPad X61 to coreboot is the kind of project that makes old hardware fans grin like kids. It keeps a beloved machine useful, strips away vendor cruft and proves once again that the repair-and-reuse crowd still has real bite.

OneDrive reminds you rent is rent

Microsoft is putting an expiry clock on some OneDrive data, which is exactly the kind of cloud fine print people dread. The promise of 'your files anywhere' keeps colliding with subscription rules, retention windows and the ugly fact that rented storage is not ownership.

AI Bills Look Brutal

One estimate claims Anthropic and OpenAI may spend more than $1,000 delivering what customers pay $100 for. That turns the AI gold rush into a very expensive magic trick, and it raises the ugly question of who eats the loss.

AI Camera Trail Goes Wrong

A Flock plate reader allegedly helped link the wrong man to a violent crime, showing how fast shaky automation can snowball into handcuffs. When surveillance tech sells certainty it cannot truly provide, ordinary people pay the price.

Breach Warnings Keep Getting Slower

After loading the 1,000th breach into Have I Been Pwned, Troy Hunt says the delay between attack and disclosure is getting worse, not better. In plain English, your data can be gone for months before anyone bothers to tell you.

Data Centers Drain Drought States

A report says U.S. data centers used 264 billion gallons of water while drought grips much of the country. The AI boom keeps demanding bigger buildings and bigger promises, but the fine print is starting to sound like water and heat.

Texas Grid Eyes Data Center Strain

Texas grid operators flagged reliability risks after some data centers and crypto sites failed voltage tests. Everyone wants endless compute, right until the lights flicker in summer and the power system reminds us it has limits.

DeepSeek Tags OpenAI Again

A benchmark write-up says DeepSeek V4 Pro beat GPT-5.5 Pro on precision, feeding the sense that the AI race is no longer a one-horse Silicon Valley parade. Cheaper challengers keep landing awkward hits where prestige used to be enough.

Designers Now Reach for Claude

One designer argued Claude has become more useful than Figma for early product thinking, sketches, and iteration. That says a lot about how fast chat tools are creeping from coding into the messy, human territory of design work.

Engineers Feel the Ground Shift

A software developer wrote that LLMs are eroding their career, not by replacing every skill, but by changing what companies value and how quickly they expect results. It reads like the quiet panic many people have been trying not to say aloud.

Trust Issues Hit AI Coding

This automated doubt workflow argues Claude Code is useful only when you keep it on a short leash, verify everything, and assume it will confidently wander off course. The mood has shifted from blind faith to strict supervision.

AI Memory Gets a Diet

A new trick called speculative KV coding promises up to 4x lossless compression of the memory used while models answer prompts. It is the kind of backstage gain that matters because better AI often comes from cheaper plumbing, not louder hype.

Google Wants Your Clicks Back

One sharp take says Google is quietly turning users into unpaid search quality raters as AI answers kill the click. If fewer people visit websites, the web loses the signals that once kept search honest and useful.

DVD Ripping Is Weirdly Easy

Something treated like a federal crime in 1999 can now be done in 2026 with a cheap drive and free software. The story is part nostalgia, part absurdity, and part reminder that media locks age badly while curiosity does not.

Gamers Fight the Shutdown Button

The Stop Killing Games campaign keeps pushing back on publishers that sell games, then pull the plug and leave buyers with nothing. It taps a growing feeling that digital ownership is too often a rental dressed up as a purchase.

YouTube Without the Brain Worms

NoSuggest strips away YouTube recommendations, autoplay, and notifications so people can watch what they chose instead of whatever the algorithm shoves next. The appeal is obvious: less slot machine, more actual video library.

Teenage Engineering Cuts Records Again

Teenage Engineering unveiled the APC-2, a professional record cutter aimed at real-time disc making. It is expensive, niche, and gloriously stubborn in the best way, a shiny reminder that some hardware still wants craft, not scale.

Privacy Phone Users Get Flagged as Suspicious

A GrapheneOS user says identity firm Yoti auto-flagged the phone and reported it to authorities just for running a privacy-focused system. That lands like a bad joke: choosing more security now looks suspicious by default.

Google Opens a Giant Checkbook for SpaceX

A filing says Google will pay SpaceX about $920 million a month for compute tied to Google Cloud and Gemini services. That number is so huge it turns the AI infrastructure race into pure stadium economics.

Steam Networking Trouble Enters Month Three

Players say Valve's peer-to-peer stack has been broken since March, hammering online matches in games like Street Fighter 6. The loud part is not just the bug, but how long a core gaming service can wobble without a clear fix.

Nvidia Draws Up a Monster Windows Chip

A reported Nvidia design for Windows PCs packs big CPU muscle, lots of shared memory, and heaps of CUDA power. It reads like the next shot in the battle to turn everyday PCs into local AI workstations.

AI Memory Wants One Shared Language

The new Universal Memory Protocol pitches a common format for how agents store and share context across tools. It is a very online dream with real appeal: stop rebuilding memory from scratch every time a model changes.

OpenAI Sells the Agent Engineer Playbook

OpenAI argues teams should wrap Codex in tight harnesses, checks, and workflows instead of treating AI like magic dust. The message is blunt and timely: agent coding is useful, but only if you box it in hard enough.

Meta's Bot Helped Hijack Instagram Accounts

Meta confirmed thousands of Instagram accounts were taken over by abusing its AI chatbot during recovery flows. It is the sort of own goal that makes every promise about safe AI automation sound a lot less calming.

Protein AI Goes Hunting for New Drugs

Biohub released an open protein world model meant to map biology and design strong binders in the lab. That gives the AI boom something more serious than chatbots: a shot at speeding up real biotech discovery.

Gigabit Internet Still Looks Like Overkill

A fresh argument says gigabit broadband is still more flex than need for most homes, even with streaming, gaming, and backups. It stings because it sounds true: providers keep selling speed while many real bottlenecks live elsewhere.

Your Cloud Memory Graph Is Probably Lying

One deep dive shows an AWS Lambda app was not leaking memory the way dashboards suggested. The real lesson is nastier and more useful: Linux memory metrics can mislead smart teams into fixing the wrong thing for days.

Warner Music Cuts Database Chunks to Seven Days

Warner Music Group's innovation team shrank TimescaleDB chunks from 30 days to 7 and saw saner performance and maintenance. It is catnip for engineers because the win came from boring schema choices, not another shiny tool.

VHS Nostalgia Gets a Surprisingly Sharp Upgrade

ntsc-rs brings crunchy analog TV and VHS artifacts to modern video with unusual care. The appeal is obvious: in an age of clean AI polish, people still want their footage to look gloriously haunted and homemade.

A Huge Free Art Library Opens Wider

The Public Domain Image Archive now offers more than 11,000 out-of-copyright works ready to browse and reuse. It is the internet behaving for once: a clean, useful pile of culture that does not ask for a subscription first.

New York slams brakes on data centers

New York lawmakers sent a bill for a one-year freeze on new large data centers, a blunt sign that AI growth is crashing into power and climate worries. The age of endless server farms just met local politics, and it may not stop here.

Cloudflare bot boom claim gets ripped apart

A fierce write-up says Cloudflare's big bot traffic jump is more stage magic than hard fact, arguing the company mixed up crawler counts and hype. With everyone waving around AI numbers lately, this takedown landed like a cold bucket of water.

Microsoft leak says AI should hook users

Leaked material around Microsoft's Scout assistant reportedly talks about making it more addictive, which is exactly the kind of phrase that turns every trust pitch into a punchline. Stop calling it help if the plan is habit-forming.

Index giant snubs SpaceX and AI stars

S&P Dow Jones reportedly kept SpaceX out of the S&P 500, a reminder that the hottest private tech names still cannot hoover up passive index money on demand. It also throws shade at easy-entry dreams for OpenAI and Anthropic.

The internet gets a plain LLM guide

One of the day's biggest hits was a clean, patient guide to how LLMs and transformers actually work. No smoke, no mysticism, just the moving parts explained in plain English. In a week full of AI chest-thumping, that felt oddly refreshing.

Claude coding panic meets the rsync receipts

The rsync blowup kept rolling as people dug into whether Claude-assisted code really added bugs or just became a handy villain. The sharper takeaway was uglier: once AI-written patches land in core tools, trust gets expensive fast.

Google shrinks Gemma for real devices

Google pushed Gemma 4 QAT models aimed at phones and laptops, pitching smaller AI that still feels capable. The message is clear enough: frontier labs do not want all the action sitting in giant data centers forever, or at least not publicly.

Sakana chases the self-improving AI dream

Sakana AI unveiled an RSI Lab focused on recursive self-improvement, which is either bold research or the kind of phrase that instantly makes people check the exits. Either way, the frontier lab race keeps drifting from tools to destiny talk.

AI bug fixers finally face real tests

A new benchmark put LLM agents up against real security flaws, and the results were messy in a useful way. Some models can help, but nobody sane should call them reliable patch machines yet. The hype train keeps meeting the brake pedal.

Microsoft puts durable jobs inside Postgres

Microsoft open-sourced pg_durable, which puts long-running, fault-tolerant jobs inside PostgreSQL instead of yet another tangle of side systems. Database fans loved the audacity; everyone else quietly wondered how many workflow tools just got insulted.

Britain drops Stripe for Dutch payments rail

The UK government picked Adyen for GOV.UK Pay, pushing out Stripe for a huge chunk of public payments. It is a reminder that government tech buying can still move markets, and that boring plumbing is where serious platform battles are won.

Popular React table package gets hit

The maintainer behind mantine-datatable was reportedly locked out after a compromise, leaving users to untangle trust, packages and damaged accounts. Another miserable reminder that open source keeps running on fragile human shoulders and shaky security.

GitHub flips wrong switch and breaks chat links

GitHub briefly nuked some Slack and Teams subscriptions after a bad feature flag change, proving once again that one tiny switch can ruin an afternoon at scale. The outage was fixed, but it was classic platform comedy with everyone else paying.

This desalination trick leaves no salty mess

Researchers unveiled a desalination method that makes drinking water from seawater without the usual toxic leftovers, and even turns the salt waste into useful materials. It sounds almost suspiciously neat, which is why people badly want it to scale.