Skip to content

Tech News

Personalized news feed, driven by a persona and implemented by a LangGraph swarm.

AI/ML / cncf.io

Filtering and reducing AI agent telemetry with the OpenTelemetry Collector

This article from the CNCF Blog discusses how to use the OpenTelemetry Collector to filter and reduce the overwhelming volume of telemetry data generated by AI coding agents, addressing storage costs and data comprehension challenges. It likely provides practical strategies for configuring the Collector to sample, aggregate, or drop unnecessary events while retaining actionable insights.

Why it matters

For engineers building and operating AI agent systems, managing telemetry volume is critical to controlling observability costs and maintaining signal clarity, and this article offers a concrete, OpenTelemetry-native approach to solve that problem.
Eight Bits of Trust; eMudhra's CA keys cracked in milliseconds
Languages / scotthelme.co.uk

Eight Bits of Trust; eMudhra's CA keys cracked in milliseconds

One of eMudhra's certificate authority private keys is the number 4. Not a number that starts with 4, or the number 4 followed by a few hundred random digits, just, 4... Another one is 56, and another is 115. I know, because I recovered all of them from their public certificates on my laptop just now, and it took less than a second. What happened On Saturday 17th October a security researcher reported to Apple, Google, Microsoft, Mozilla and eMudhra that the private keys for eight eMudhra intermediate certificates were compromised. Chrome pushed a CRLSet update blocking them that evening and…
Left alone with nothing to do, AI agents started playing
AI/ML / techcrunch.com

Left alone with nothing to do, AI agents started playing

MIT researchers gave 13 Eko agents (Claude Opus 4.7) 30 hours of unstructured play on a simulated desert island, where they invented challenges like ring-throwing and rock-stacking. This free exploration led to emergent skills (e.g., throwing blocks farther mid-jump) and later outperformed non-playing agents on specific tasks, suggesting goal-free exploration as a training paradigm.
Two-panel illustration: on a green background, two hand-drawn geometry diagrams of lettered circles with red-shaded shapes; on a red background, scattered grey dots with a single green dot among them.
AI/ML / anthropic.com

LLMs are making formal software verification feasible at a much larger scale

Chapter 01 : Introduction Keane Lucas, Aengus Lynch, Luna Tong, Luke Champine, Zygi Straznickas, Dino Distefano, Harshikaa Agrawal Evyatar Ben Asher, Ava Howell, Sami Jawhar, Chris Rohlf, Nickolai Zeldovich Nicholas Carlini, Rajashree Agrawal, Jason Gross Most software is riddled with bugs, which AI can now easily find and exploit . The standard defenses against cyberattacks are no longer sufficient: we need ways to prove that code is free of bugs. Formal software verification could provide this guarantee, but it has historically required years of expert effort, making it impractical for most…
Who on-calls an AI? This is the unresolved crisis of agent ownership
AI/ML / thenewstack.io

Who on-calls an AI? This is the unresolved crisis of agent ownership

At 2:07 a.m., an AI agent starts failing.  The application is still running, but requests are timing out because a The post Who on-calls an AI? This is the unresolved crisis of agent ownership appeared first on The New Stack .
The Pulse: Anthropic builds its own source control system
Open Source / blog.pragmaticengineer.com

The Pulse: Anthropic builds its own source control system

Git had too many performance problems at 30,000 PRs per day, so Anthropic built its own. I talked with the team, including asking why they didn’t open source its new system, ‘Scion.’
A CA private key was the number 4. Your curl still trusts it (CVE-2026-111943)
Languages / dev.to

A CA private key was the number 4. Your curl still trusts it (CVE-2026-111943)

Eight eMudhra intermediate CA keys were mis-generated — recoverable from the public certificates since April 2025. I recovered one in 4 guesses. Browsers are safe; your curl, git and CI may not be.
Debunking "Faster than Rust": Malicious Benchmarks in Tov
Languages / blog.fyralabs.com

Debunking "Faster than Rust": Malicious Benchmarks in Tov

A blog post debunks Tov, a language marketed as 'faster than Rust' that transpiles TypeScript to C via Clang. The author argues the claim is logically impossible (C vs. C) and proves benchmarks for `bigint_fib`, `dispatch`, and `map_insert` were maliciously crafted to slow Rust down. Corrected benchmarks show Rust has equivalent or better performance.
Presentation: Actors for Agents: Patterns for Production AI
AI/ML / infoq.com

Presentation: Actors for Agents: Patterns for Production AI

Manju Rajashekhar (Mad Labs, ex-Twitter/Etsy) applies the actor model to production AI agents, tackling runtime challenges like state ownership, write-through state, and supervision boundaries. This distributed systems pattern manages failures, continuations, and mutable execution graphs, moving agentic workflows beyond fragile demos to resilient architectures. The talk provides a concrete architectural playbook for reliable, long-horizon agent execution with strong governance.
OpenAI donates Open Responses, its spec for a common LLM API, to the Agentic AI Foundation
AI/ML / thenewstack.io

OpenAI donates Open Responses, its spec for a common LLM API, to the Agentic AI Foundation

OpenAI on Thursday announced that it is donating the Open Responses API specification to the open source Agentic AI Foundation The post OpenAI donates Open Responses, its spec for a common LLM API, to the Agentic AI Foundation appeared first on The New Stack .