A tech digest, every weekday. About
212 items across 29 sources in the last seven days. The email carried the 37 that mattered most.
Automatic summaries were unavailable for this edition; a classifier chose and ranked the stories.
Google Unveils Gemini 4 Argon, Retaking Benchmark Lead Over OpenAI and Anthropic
Google has unveiled Gemini 4 Argon, a new frontier AI model that it says leads or ties rivals on 13 of 18 disclosed benchmarks. The model is initially being released only to trusted cyber defenders and select pre-release testers, with broader availability planned later. VentureBeat reports: Google is not claiming that Gemini 4 Argon wins every benchmark. But across the benchmark table disclosed in
The Internet has a second audience
More than half the traffic reaching sites on Cloudflare is now automated, and AI agents are the fastest-growing part of it. We're giving site owners the tools to see who's visiting, decide who gets in, and charge for access.
[AINews] Gemini 4 Argon: GDM’s answer to Astra/Fable, with 1M output
... but you can’t try it yet unless you are “government users and trusted cyber defenders in the Fairwind Program”
AI & LLMs
Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agents
Monetization Gateway beta: charge AI agents for consumption with HTTP 402
Cloudflare’s AI Gateway, Ceramic.ai, Stocktwits, and more are using the Cloudflare Monetization Gateway today to charge agents for access to tokens, APIs, and MCP tools. U.S.-based sellers can now apply for access to the closed beta.
Disrupting a coordinated model-distillation campaign
Learn how OpenAI disrupted a campaign to extract protected model reasoning and is strengthening defenses against adversarial distillation.
Quoting Matthew Green
[...] Put these pieces together and you have the two halves of a worm: a payload that hijacks the agent, and an agent that will carry the payload to the next agent. Agents in separately-isolated sandboxes discovered that they could leave instructions for each other in a shared package cache, and those instructions changed what the recipients did. Replace the package cache with email, Slack and sha
Why Dwarkesh is Wrong about Computer Use + How OpenAI shipped its Jev competitor in 1 Week
Our DevDay coverage - the first pod on the DevDay lineup - dives in with the leaders of OpenAI’s CUA team and API platform.
Cut your AI spend with AI Gateway's Auto Router
Cloudflare AI Gateway now features a model router that evaluates request complexity using an edge-deployed classifier to select the optimal model. By balancing expected output quality against token costs, organizations can dramatically cut AI spend while maintaining performance.
Pay Per Use: when AI uses your work, you should get paid
Pay Per Use is now in beta. AI companies report when they use publishers’ content, and Cloudflare handles billing, payouts, and reporting, so our customers are paid according to use.
Open TTS Leaderboard: Scalable Evaluation for Multilingual Text-to-Speech and Voice Cloning
Web & Frontend
Vercel CDN no longer caches responses with Vary: Cookie
Vercel CDN no longer caches origin responses when Vary includes Cookie. Vary tells a cache which request headers may change the response. Because cookies often contain visitor-specific values, varying by Cookie can create many cache entries that are rarely reused. The response is still served normally but isn't stored for future requests.You can identify these responses by checking the x-vercel-ca
Devtools & Platform
5x faster Edge Functions: V8 isolates to Firecracker MicroVMs
Detect and send production issues straight to your agent
You can now use built-in error monitoring in Cloudflare Workers to group production failures and send stack traces, logs, traces, and application context directly to a coding agent to investigate further and open a pull request.
Vercel Agent now installs private packages from npm and custom registries
Vercel Agent can install private dependencies from npm and custom registries using credentials stored as shared environment variables on Vercel. npm, pnpm, and classic Yarn running in Agent sessions authenticate as they do in Vercel builds.To get started, add a shared environment variable for Development or Preview:Use NPM_TOKEN for private packages hosted on registry.npmjs.org.Use NPM_RC to confi
Opt-in dist-tag permissions for npm trusted publishing
Trusted publishing configurations for npm can now be granted permission to manage dist-tags (e.g., promoting a version to latest, updating next and beta pointers) using short-lived OIDC credentials instead of… The post Opt-in dist-tag permissions for npm trusted publishing appeared first on The GitHub Blog.
EDG C++ front-end goes public
Vercel Sandbox now supports Secure Compute
Vercel Sandbox now supports Secure Compute, connecting sandboxes to a team's dedicated network. Public-internet traffic exits through the network's static IPs, and sandboxes can reach private resources in an AWS VPC through VPC peering.You can attach a sandbox to your network in your code or via the CLI. In your code, pass an existing Secure Compute network ID when creating a sandbox:In the CLI, a
X25519-only TLS ends for GHE.com on October 7
Beginning October 7, 2026, GitHub Enterprise Cloud with data residency will no longer accept TLS connections from clients that offer only X25519 for key agreement. Most customers don’t need to… The post X25519-only TLS ends for GHE.com on October 7 appeared first on The GitHub Blog.
GitHub Advanced Security trials for GitHub Team
GitHub Team customers can now start self-serve trials of GitHub Advanced Security to evaluate GitHub Code Security and GitHub Secret Protection. Start a trial from your organization’s Overview page, Billing… The post GitHub Advanced Security trials for GitHub Team appeared first on The GitHub Blog.
Deep Dives
Cloudflare Containers, rebuilt to scale agent sandboxes
Cloudflare Containers now start 6x faster, let your agent choose each sandbox's image and instance type at runtime, and support filesystem snapshots in public beta, all controlled from a Durable Object.
What TLA+ can and can't check
Branch Target Reuse: Spectre-v2 Attacks in JIT Engines
Comments
How we built an async-aware Python profiler
Learn how we made Python profiling async-aware while reducing profiler overhead and preserving context across asyncio tasks.
How Etsy gets its mobile apps ready for peak traffic
How Etsy's mobile team preps for peak traffic: a three-phase code slush, a break-glass alert list, surface-level tagging, and what a crash costs in GMS.
We used a database as a message queue. Now we use Kafka
Comments
Quick Hits
Responsible Release of AI-Generated Mathematics
Major rsync upgrade in Debian because of 33 CVEs
apt-listchanges --which=both -f text --since=3.4.1+ds1-5+deb13u4 /var/cache/apt/archives/rsync_3.5.0+ds1-0+deb13u1_amd64.deb apt-listchanges: Reading changelogs... apt-listchanges: News rsync (3.5.0+ds1-0+deb13u1) trixie-security; urgency=medium In order to fix 33 CVEs, I have decided to bump the package to 3.5.0 rather than backporting all patches individually. After analysing the extra changes
Reddit Is Killing RSS Feeds, Ending Public API Access
An anonymous reader quotes a report from TechCrunch: Amid a number of updates for moderators and developers announced Wednesday comes bad news for supporters of a more open web: Reddit is ending support for RSS feeds. Reddit says RSS has now become a "common surface for large-scale scraping and automated abuse," which is why it's made the decision to wind down RSS feeds on its platform. "We know R
Show HN: Ledge.sh – Runnable Markdown Notes
Hackers Stole Millions of US Military Personnel Records During Months-Long Data Breach
A months-long breach of the Defense Manpower Data Center exposed personal information belonging to roughly 2.8 million living current and former U.S. military personnel and staff, plus records for nearly 300,000 deceased people. Attackers reportedly exploited a file-sharing vulnerability between October 2025 and July 2026, accessing unencrypted records that included Social Security numbers, dates
ESXi Exploitation in the Wild
Comments
Singapore govt dating app uses Gale-Shapley stable marriage algorithm
Halfspace experimental IDE for solid modeling with distance fields
From the timeline
Stop saying they are using it wrong.
Stop saying they are using it wrong. Listen without defense. @marktechson.com on the two feedback paths you can actually act on. Full episode and notes: https://www.epicproduct.engineer/lead-with-empathy-and-listen-without-defense-product-sense-with-mark-techson~5z61q
A distributed system is one in which the failure of a computer you did not even know existed can render your own computer unusable.Leslie Lamport