Everything else

The full seven-day window, including what did not make the email.

• Mario Meets ParetoHacker News · 8/6 · 1028 points
• AMD acquires Taalas to boost inference performance by etching models in siliconHacker News · 8/6 · 625 points · in the email
• Taste Is All That's LeftHacker News · 8/6 · 393 points
• GitHub Actions and Pages are experiencing degraded availabilityHacker News · 8/6 · 391 points
• Humans missed 1 in 3 threats approving AI agent commands across 40k game runsHacker News · 8/6 · 293 points · in the email
• Quake – 30th Anniversary UpdateHacker News · 8/6 · 288 points
• Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free usersHacker News · 8/6 · 227 points · in the email
• Scientists discover Kelvin-Helmholtz Instability on the surface of the SunHacker News · 8/5 · 213 points
• New Mexico court orders Meta to pay $567m over harms to children’s mental healthHacker News · 8/7 · 207 points
• Launch HN: ProvenMetal (YC S26) delivers circuit boards in days instead of weeksHacker News · 8/6 · 202 points
• Herdr is joining Y Combinator. The runtime stays openHacker News · 8/6 · 197 points
• I stopped trusting USB-C cable labels and started testing themHacker News · 8/3 · 169 points
• Welcoming the Nepalese Government to Have I Been PwnedHacker News · 8/6 · 145 points
• My phone detects going on a run as “someone snatching my phone and running off”Hacker News · 8/6 · 134 points
• The simple elegance of the integrated timing belt loopback fastenerHacker News · 8/2 · 116 points
• Sao Paulo resident transforms degraded area into urban forestHacker News · 8/1 · 115 points
• Bioengineered chewing gum may offer a way to fight HPV and other microbesHacker News · 8/6 · 113 points
• Show HN: The Channels SDK – Bring Any Agent to Any Channel (Slack, MS Teams)Hacker News · 8/6 · 102 points
• Inside vLLM: Anatomy of a High-Throughput LLM Inference System (2025)Hacker News · 8/6 · 96 points · in the email
• Dress Made of Living Mycelium Can Renew and Repair ItselfSlashdot · 8/7
• Scientists Make First Viruses Designed By AISlashdot · 8/7
• Court Orders Meta To Establish $567 Million Fund To Abate Harms To YouthSlashdot · 8/7
• Google's $15 Billion India Data Center Project Battles Water, Wildlife ConcernsSlashdot · 8/6
• 'Tower Dump' Warrants Ruled UnconstitutionalSlashdot · 8/6
• Azure CTO Pastes Doom Into Paint One Frame At a TimeSlashdot · 8/6
• OpenAI's Models Shared Hacking Tips On a Secret Messaging Board Before Hugging Face BreachSlashdot · 8/6 · in the email
• The US Government's Mario, Pokemon and Naruto Meme Posting Could Damage These Franchises, Japanese Officials WarnSlashdot · 8/6
• FCC Kills TV Ownership Cap, Claiming Authority Over Limit Set By CongressSlashdot · 8/6
• Times Magazine Now Serves Ads That Only AI Chatbots Can SeeSlashdot · 8/6
• Meta AI Hacked External Systems During Cybersecurity TestingSlashdot · 8/6
• Broadcaster Wins Broad US Blocking Injunction Covering Pirate Sites That Don't Exist YetSlashdot · 8/6
• FDA Approves First mRNA Flu ShotSlashdot · 8/6
• BMW Blasts Its Cars' Internal Screens With Aggressive AdsSlashdot · 8/6
• New Images of the Sun Show Its Surface In the Finest Detail YetSlashdot · 8/6
• tl;dv (Too Lazy; Didn't Validate): 181,874 Meetings Left Wide OpenLobsters · 8/6 · in the email
• Zig's Io.Threaded is NeatLobsters · 8/6 · in the email
• A shell exclamation mark is not for yelling. Be lazyLobsters · 8/7
• jujutsu 0.44.0Lobsters · 8/6 · in the email
• i wonder if you could make a useful DOS executable with just emojiLobsters · 8/6
• I got a 1998 CD-ROM world atlas running on Windows 11Lobsters · 8/7
• Taste Is All That's LeftLobsters · 8/6
• Super Mario DerivationsLobsters · 8/6
• the web server deployment model breaks at hobby scaleLobsters · 8/6 · in the email
• Setup a simple web server with bozohttpd on NetBSDLobsters · 8/6
• python string literals are kinda funnyLobsters · 8/6
• I'm leaving OpenAI to build Jurassic ParkLobsters · 8/7
• How to Make a Nintendo 64 Game in 2026Lobsters · 8/6
• The following is a valid DOS COM executableLobsters · 8/6
• AI-generated vulnerability patches require human reviewLobsters · 8/6 · in the email
• celld — Durable Objects, self-hostedLobsters · 8/6
• cl-forth: Forth 2012 Standard in CLLobsters · 8/6
• schrodingers-toctou: The binary you run is not the program you wroteLobsters · 8/6
• Introducing Kitesurf: The agent-first browser that runs in V8 isolates on Cloudflare WorkersLobsters · 8/7 · in the email
• Crubit, C++/Rust Bidirectional Interop ToolLobsters · 8/6
• Zapscape - Guest to host escape in KVM/x86Lobsters · 8/6
• Offensive Internet PostureLobsters · 8/5
• A Vision for CargoLobsters · 8/5
• Finally adding recursive functions to FutharkLobsters · 8/6
• rust-lang/rust is adopting an LLM policyLobsters · 8/5 · in the email
• iPhone 18 Pro and Foldable iPhone Could Sell Out Fast Due to Memory ShortageMacRumors · 8/7
• Original iPhone Loses Its Last Cellular Network as T-Mobile Kills 2GMacRumors · 8/6
• Google Maps Can Now Order Your Dinner and Track Your Train DelaysMacRumors · 8/6
• Madden NFL 27 Arcade Edition Now Available on Apple ArcadeMacRumors · 8/6
• BMW Owners Angry Over Spider-Man Ads Appearing on Dashboard ScreensMacRumors · 8/6
• OpenAI's ChatGPT Speaker Will Be Hockey Puck-Sized and Cost Over $300MacRumors · 8/6
• Free ChatGPT Users Get Unlimited Text Chats and GPT-5.6 LunaMacRumors · 8/6 · in the email
• Apple Releases macOS Tahoe 26.6.1 With Security FixesMacRumors · 8/6
• Last Day for Up to $350 Amazon Gift Card With Samsung Galaxy Pre-OrdersMacRumors · 8/6
• Apple Raises Trade-In Values for iPhone, Mac, and MoreMacRumors · 8/6
• iPhone 18 Pro Launching Next Month With These 12 New FeaturesMacRumors · 8/6
• Apple Scrambling to Secure Memory Ahead of iPhone 18 Pro LaunchMacRumors · 8/6
• Amazon Takes Up to $190 Off M5 MacBook Air With Best Deals Since Price HikeMacRumors · 8/6
• Apple Account Card in Wallet Now Available in Four More CountriesMacRumors · 8/6
• AirPods Pro 3 Drop to One of the Best Prices of the Year So Far at $189.99MacRumors · 8/6
• Apple's Bid to Use Chinese Chipmaker as Leverage BackfiresMacRumors · 8/6
• OpenAI Asks Judge to Dismiss Apple's Trade Secrets LawsuitMacRumors · 8/6
• Meta's New Mac Coding Agent Costs Up to 20x Less If You Let Meta Train on Your DataMacRumors · 8/6 · in the email
• Disney+ Could Get a Free Ad-Supported TierMacRumors · 8/5
• SpaceX Plans Starlink Mobile Service to Compete With Verizon, AT&T and T-MobileMacRumors · 8/5
• Reimplementing htmx from scratchJavaScript Weekly · 8/4
• Node.js 26.7, vlt 1.0, and Shai-Hulud ReturnsNode Weekly · 8/6 · in the email
• Mariano-Florentino (Tino) Cuéllar to join Anthropic as Chief Global Affairs OfficerAnthropic · 8/4
• Improving Fable 5's biology safeguardsAnthropic · 8/7
• Improving GPT‑5.6 Sol in ChatGPT—and expanding access to GPT-5.6 Luna for free usersOpenAI · 8/6 · in the email
• Working with the American Psychological Association on youth mental health and AIOpenAI · 8/6
• From asking to doing: How the world is putting ChatGPT to workOpenAI · 8/6
• Third-party cyber evaluations involving OpenAI modelsOpenAI · 8/4
• New ways to learn and teach with ChatGPT Work and CodexOpenAI · 8/4
• Apple is getting this wrongOpenAI · 8/3
• How we built a realtime system for responsive voice AI in six monthsOpenAI · 8/3
• Circles powers telco personalization with OpenAI technologyOpenAI · 8/3
• Ten advances in mathematics and theoretical computer scienceOpenAI · 8/1
• Advancing responsible AI across EuropeOpenAI · 7/31
• Building abundant intelligenceOpenAI · 7/31
• WeatherNext: AI model achieves breakthrough in forecasting cyclonesGoogle DeepMind · 8/6
• Baseten on Hugging Face Inference Providers 🔥Hugging Face · 8/6
• Deploy local agents everywhere with LFM2.5-2.6BHugging Face · 8/4
• datasette 1.0a38Simon Willison · 8/6
• datasette 0.65.3Simon Willison · 8/6
• Simon Willison on Technical BloggingSimon Willison · 8/6
• An AI model from Meta also hacked another company during testingSimon Willison · 8/6
• Introducing Muse Code and Muse Spark 1.2Simon Willison · 8/5 · in the email
• Third-party cyber evaluations involving OpenAI modelsSimon Willison · 8/5 · in the email
• Incident Report: unsanctioned agent behaviour during cyber testingSimon Willison · 8/5 · in the email
• One-shotting a Raccoon Heist game using Claude Fable 5Simon Willison · 8/5
• New release of LLM adds support for reasoning traces, OpenAI Responses, server-side tools, and smarter loggingSimon Willison · 8/4
• llm-anthropic 0.26Simon Willison · 8/4
• PipeNetwork/minimax-h3-mlxSimon Willison · 8/4
• llm 0.32Simon Willison · 8/4
• Quoting Steve YeggeSimon Willison · 8/4
• Don't be a meat proxySimon Willison · 8/3
• Quoting David Crawshaw's promptSimon Willison · 8/3
• Devtools must be open source (exe.dev)Simon Willison · 8/3
• condense-json 1.1Simon Willison · 8/3
• condense-json 1.0Simon Willison · 8/2
• Open letters about AI developmentSimon Willison · 8/2
• July 2026 newsletterSimon Willison · 8/2
• Quoting Greg BrockmanSimon Willison · 8/1
• datasette-apps 0.2a0Simon Willison · 8/1
• Ten advances in mathematics and theoretical computer scienceSimon Willison · 8/1
• deepseek-ai/DeepSeek-V4-Flash-0731Simon Willison · 7/31
• Stateless MCP has recaptured my interest (and inspired mcp-explorer and datasette-mcp)Simon Willison · 7/31
• llm-mcp-client 0.1a0Simon Willison · 7/31
• Oxide and Friends: The Open Weight Revolution with Simon WillisonSimon Willison · 7/31
• smevals - a small eval suite for evaluating models, prompts, and harnessesSimon Willison · 7/31
• Slack Emoji MakerSimon Willison · 7/31
• datasette-agent 0.4a0Simon Willison · 7/31
• [AINews] AMD buys TaalasLatent Space · 8/7 · in the email
• [AINews] Jeff, Sanjay, Oriol, and Quoc depart DeepMind; Demis to Chair; Koray to SVP — what is going on at GDM???Latent Space · 8/6
• [AINews] Megakernels are so dead and so backLatent Space · 8/5
• Unpacking ChatGPT Work: the Agent for a Billion UsersLatent Space · 8/4
• [AINews] Qwen 3.8 Max(2.4T) and 27B, new open weights models for Coding and CoworkLatent Space · 8/4
• The Inference Engineering Masterclass — Philip Kiely & Ali Taha, BasetenLatent Space · 8/3
• [AINews] not much happened todayLatent Space · 8/1
• Modeling Device Capabilities for AnalyticsNetflix Tech · 7/31
• Cloudflare AI Search: give your agents a search engine for your dataCloudflare · 8/6
• The next generation of MCPCloudflare · 8/6 · in the email
• From ranking to recommended: get your site ready to thrive in the age of AI agentsCloudflare · 8/6
• Building an open Agentic Internet: readable, discoverable, callable, and payableCloudflare · 8/6
• Introducing Kitesurf: The agent-first browser that runs in V8 isolates on Cloudflare WorkersCloudflare · 8/6 · in the email
• Give any website a WebMCP interfaceCloudflare · 8/6 · in the email
• Cloudflare is the only vendor named a Visionary in 2026 SASE and SSE reportsCloudflare · 8/5
• The Agent Access ModelCloudflare · 8/5
• How we’re rethinking work at Cloudflare with Cloudflare OSCloudflare · 8/5 · in the email
• Cloudflare OS: an open platform for agents, apps, and workCloudflare · 8/5 · in the email
• WriteGuard: fine-grained controls for MCP ServersCloudflare · 8/5
• Catching rogue AI behavior with identity-aware analyticsCloudflare · 8/5
• The Agent Development Lifecycle has arrived on CloudflareCloudflare · 8/4
• Announcing Cloudflare Wallets: the programmable wallet for the agentic InternetCloudflare · 8/4
• Run CI/CD for millions of repos — on your platform, on CloudflareCloudflare · 8/4
• How Cloudflare enforces engineering standards using AICloudflare · 8/4
• Introducing: Cloudflare AgentsCloudflare · 8/4
• Your agent can now debug Workers with local tracingCloudflare · 8/4
• How we built a software factory to drive Astro’s GitHub issue count to zeroCloudflare · 8/4
• Your agent needs a computer, not a container — introducing @cloudflare/computerCloudflare · 8/3
• From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta’s Ads RankingMeta Engineering · 8/5
• GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation ModelMeta Engineering · 8/3
• A guide to slash commands in the GitHub Copilot appGitHub Blog · 8/6
• How we took malware advisories beyond npmGitHub Blog · 8/6 · in the email
• How the GitHub legal team used Copilot CLI to streamline their workflowsGitHub Blog · 8/4
• Turn one giant AI-generated pull request to a reviewable stackGitHub Blog · 8/4
• Don’t stop early: Case-folding source code at memory speedGitHub Blog · 7/31
• Kimi K3 is now available in GitHub CopilotGitHub Changelog · 8/6
• Customize code scanning default setup at scaleGitHub Changelog · 8/4
• Retiring the Copilot Billing Preview appGitHub Changelog · 8/4
• Upcoming deprecation of GitHub Spark on github.comGitHub Changelog · 8/4
• Customize Dependabot pull request branch namesGitHub Changelog · 8/4
• Code coverage automatic enablement in Code Quality settingsGitHub Changelog · 8/4
• CodeQL 2.26.2 adds Swift 6.3.3 and Kotlin 2.4.10 supportGitHub Changelog · 8/4
• Customize the reasoning level for Copilot cloud agentGitHub Changelog · 8/3
• Enterprise team specialization for managed settingsGitHub Changelog · 8/3
• Trigger Copilot automations with commentsGitHub Changelog · 8/3
• Introducing Agent PluginsVercel · 8/6 · in the email
• Seedance 2.5 now available on Vercel AI GatewayVercel · 8/6
• Marketplace integrations now install provider skillsVercel · 8/6
• Ling 3.0 Tiny is now available on AI GatewayVercel · 8/6
• Introducing Agent Plugins 1.0.0Vercel · 8/6 · in the email
• Pause workflows for approval with Chat SDKVercel · 8/6 · in the email
• Set your own project avatarsVercel · 8/5
• New setup page after domain checkoutVercel · 8/5
• Export AI Gateway traces with Vercel DrainsVercel · 8/5
• Introducing the new v0 APIVercel · 8/5 · in the email
• AI Gateway is now available on AWS MarketplaceVercel · 8/5
• Muse Spark 1.2 is now available on Vercel AI GatewayVercel · 8/5 · in the email
• Vercel Sandbox now supports 10,000 concurrent sandboxes and 5,000 vCPUs per minuteVercel · 8/5
• Search and buy domains in the dashboardVercel · 8/5
• Measure time between steps in Vercel WorkflowsVercel · 8/5
• Full Sandbox egress firewall now available on Hobby planVercel · 8/5
• Vercel Sandbox now supports Devin OutpostsVercel · 8/4
• Next.js 16.3 support on VercelVercel · 8/4
• Skill packs are now available on skills.shVercel · 8/4
• Deploys are now up to 33% faster for apps with many ISR pagesVercel · 8/4
• Give your eve agent a browserVercel · 8/4
• DeepSeek V4 Flash is 90% off through Novita on AI GatewayVercel · 8/4
• Vercel WAF for Blob is now generally availableVercel · 8/3
• How Factory scaled its cloud backend to one billion monthly requests on VercelVercel · 8/3
• Share Vercel Container Registry repositories across teamsVercel · 8/3
• Qwen 3.8 Max now available on Vercel AI GatewayVercel · 8/2
• AI Gateway now supports team and project spend budgetsVercel · 7/31
• Discord Patch Notes: August 4, 2026Discord Engineering · 8/6
• General Availability of Mobile Platform Support Arrives in Discord Social SDK Version 1.10Discord Engineering · 8/6
• Squircles, Styles, and Spacing: How Your Feedback is Helping Improve MobileDiscord Engineering · 8/4
• How we built an automated debugging workflow at SentrySentry · 8/6
• Your OTel spans, our errors: A Sentry love story in one traceSentry · 8/5
• Node.js 26.7.0 (Current)Node.js · 8/5 · in the email
• Node.js 26.6.0 (Current)Node.js · 8/3
• Node.js 24.19.0 (LTS)Node.js · 8/3