Anthropic Pauses Token-Based Billing for the Claude Agent SDK
Ars Technica AI41 days agoBusinessAnthropic has paused a planned shift to token-based billing for its Claude Agent SDK, originally scheduled to take effect Monday. The change would have significantly increased costs for power users running high-volume agentic pipelines, where token consumption far exceeds standard chat completions. By framing the decision as a pause rather than a cancellation, Anthropic signals the billing direction has not changed — only the timeline — leaving developers without a clear revised rollout date.
What Are Git Worktrees, and Why Should You Use Them?
GitHub Blog41 days agoTutorialGit worktrees let developers check out multiple branches simultaneously in separate directories, all sharing one repository's object store. The feature has existed since Git 2.5 in 2015 but has surged in relevance alongside AI coding agents that benefit from isolated working directories. This GitHub Blog tutorial covers what worktrees are, how to use them, and when they make sense.
Android 17 Launches with New Multitasking Tools as Google Expands Gemini Features
TechCrunch AI41 days agoReleaseGoogle has officially launched Android 17 alongside Wear OS 7, bringing new multitasking capabilities, enhanced parental controls, and security improvements to Android devices. The release is paired with a Pixel Drop update that deploys Google's latest AI models directly to Pixel hardware. Gemini features are notably expanded as part of this platform update.
GPT-NL: A Sovereign Language Model for the Netherlands
Hacker News (AI keywords)41 days agoBusinessGPT-NL is a sovereign large language model initiative led by TNO, the Dutch national research organization, aimed at giving the Netherlands independent AI capabilities free from reliance on foreign — primarily American — cloud providers. The project focuses on Dutch-language performance and domestic data hosting, targeting government, healthcare, legal, and enterprise use cases subject to strict data residency rules. It reflects a broader European push to build competitive, nationally controlled AI infrastructure.
10Gb/s Ethernet: Switching to a Broadcom SFP+ Module
Hacker News (AI keywords)41 days agoHardwareThis post documents the decision to move a home or small-office network to 10Gb/s Ethernet using a Broadcom-based SFP+ module. SFP+ is the dominant small form-factor standard for high-speed fiber and direct-attach copper links in both enterprise and prosumer settings. The article appears to cover hardware selection rationale and the practical compatibility considerations involved in reaching 10Gb/s throughput in a home-lab context.
Claude: Elevated Errors Across Many Models
Hacker News (AI keywords)41 days agoIncidentAnthropic posted an incident notice on its status page indicating elevated error rates across many Claude models. The incident affects API consumers relying on Claude for production workloads, automation pipelines, and developer tooling. No body content was provided beyond the incident title; users should monitor status.claude.com for updates and resolution times.
Apple 2027 Rumors: AirPods With Cameras for AI and a Second Folding iPhone
The Verge AI41 days agoHardwareBloomberg's Mark Gurman reports that camera-equipped AirPods remain on schedule for a late 2027 release, with AI integration as the stated core use case. The update, published shortly after WWDC 2026, also references a second-generation folding iPhone as part of Apple's longer-term hardware roadmap. While specific feature details are sparse, the report signals active development progress on both devices rather than early-stage conceptual speculation.
Qualcomm's Snapdragon Reality Elite Chip Targets Next-Gen Smart Glasses
The Verge AI41 days agoHardwareQualcomm has announced the Snapdragon Reality Elite, a new chip designed to power the next generation of extended-reality smart glasses and wearables. The announcement came at Augmented World Expo 2026, though a device running the chip had reportedly already been previewed in a hands-on setting before the official reveal. The move underscores Qualcomm's continued push to build the silicon foundation for an emerging smart-glasses market.
Stop Using JWTs
Hacker News (AI keywords)41 days agoOpinionA pointed opinion piece challenges the widespread adoption of JSON Web Tokens (JWTs) for session management, a practice that has become near-default in modern web development. The author argues that JWTs introduce unnecessary complexity and security pitfalls that traditional server-side sessions do not. Developers are encouraged to reconsider default technology choices and reach for simpler, more auditable authentication patterns.
TIL: Make HTTP Requests Without curl Using Bash /dev/TCP
Hacker News (AI keywords)41 days agoTutorialA developer shares a lesser-known Bash capability: the /dev/tcp virtual filesystem, which opens raw TCP connections using only shell built-ins. By writing HTTP headers to a file descriptor and reading the response back, you can fetch web resources without curl, wget, or any installed HTTP client. This technique is especially valuable in minimal containers, restricted shells, or embedded environments where external binaries may not exist.
SpaceX to Acquire AI Coding Tool Cursor to Challenge Anthropic and OpenAI★ 73
Ars Technica AI41 days agoBusinessSpaceX has announced plans to acquire Cursor, the widely used AI-powered code editor, in a bid to challenge Anthropic and OpenAI in the developer-tools market. Neither company could independently compete with the AI giants, but together they believe a combined entity can mount a credible challenge. The deal marks SpaceX's push beyond aerospace into commercial AI software, where control of developer tooling carries significant strategic and market leverage.
datasette-tailscale 0.1a0
Simon Willison's Weblog41 days agoReleaseSimon Willison released datasette-tailscale 0.1a0, an experimental alpha plugin that connects a local Datasette server to a Tailscale private network (Tailnet) via a single CLI command. Using a Tailscale auth key and a hostname flag, it starts a localhost server alongside a Tailscale sidecar, making the database reachable at a friendly Tailnet URL. The plugin relies on Python bindings for the experimental tailscale-rs Rust library; Willison has filed an upstream issue about the proxy mechanism, indicating the implementation may change.
Georgi Gerganov on Using Qwen3.6-27B Daily for Local Coding Tasks
Simon Willison's Weblog41 days agoCommentaryGeorgi Gerganov, creator of llama.cpp, endorses Qwen3.6-27B as a capable local model for everyday coding assistance, citing six weeks of daily use on Apple M2 Ultra and NVIDIA RTX 5090 hardware. He runs a minimal setup — the pi agent with `pi -nc --offline` and a short custom system prompt — for routine maintainer tasks at ggml-org. His primary constraint is PR review time, which limits how heavily he can leverage the model.
Securing the Future of AI Agents
Google DeepMind Blog41 days agoCommentaryGoogle DeepMind has published a framework called the AI Control Roadmap aimed at securing internal systems that run AI agents. The approach pairs conventional security safeguards — such as access controls and least-privilege principles — with real-time behavioral monitoring designed for the speed and autonomy of AI agents. The roadmap signals DeepMind's view that neither purely traditional nor purely AI-specific security measures are sufficient on their own.
But Yak Shaving Is Fun
Hacker News (AI keywords)41 days agoOpinionA developer blogger argues that yak shaving — the recursive chain of side tasks that emerges when pursuing any goal — is not a productivity failure to avoid, but a source of genuine satisfaction. The essay pushes back against conventional advice to prune exploratory detours, contending these tangents are where deep learning actually happens. As AI coding assistants increasingly automate such rabbit holes, the piece implicitly raises the question of what developers lose when exploration is delegated to machines.
Frontier Post-Training Recipe Review with Finbarr Timbers
Interconnects (Nathan L.)42 days agoCommentaryIn the 18th installment of his interview series, Interconnects author Nathan Lambert speaks with Finbarr Timbers about the post-training techniques used at frontier AI labs. The conversation examines the methodologies — including supervised fine-tuning, reinforcement learning from human feedback, and preference optimization — that shape model behavior after pretraining. The discussion offers a practitioner's perspective on the evolving landscape of alignment and capability tuning at scale.
Qwen-Robot Suite: A Foundation Model Suite for Physical World Intelligence
Hacker News (AI keywords)42 days agoReleaseAlibaba's Qwen team has announced Qwen-Robot Suite, a suite of foundation models targeting physical world intelligence — AI systems that reason about and interact with real environments. The release expands the Qwen ecosystem beyond language and vision into embodied and robotic AI, a domain demanding integrated perception, spatial reasoning, and physical action generation. The suite format suggests multiple specialized components, potentially suited to manipulation, locomotion, and instruction-following tasks in robotic deployments.
Probably Raises $9M to Build More Reliable AI
TechCrunch AI42 days agoBusinessProbably, an AI reliability startup, has raised $9 million in funding to tackle one of the field's most persistent problems: hallucinations and factual errors in AI outputs. The company's stated goal is to prevent inaccurate information from ever reaching end users, targeting accuracy levels comparable to traditional deterministic software. This positions Probably squarely in the growing space of AI output verification and trust infrastructure.
Cloudflare DMARC Management is Now Generally Available
Cloudflare Blog42 days agoReleaseCloudflare has moved its DMARC Management product to general availability, making it free for all customers regardless of plan tier. The tool centralizes email authentication visibility, offering deeper DMARC aggregate reporting, DNS record analysis, and SPF configuration audits within the existing Cloudflare dashboard. It gives organizations a structured path to achieve full DMARC enforcement without requiring separate third-party tooling.
SpaceX Acquires Cursor for $60 Billion in Enterprise AI Push★ 78
The Verge AI42 days agoBusinessSpaceX has announced a $60 billion acquisition of Cursor, the AI-powered code editor, days after its own IPO. The deal is framed as a strategic move to attract enterprise customers and narrow the competitive gap with Anthropic and OpenAI. The takeover was signaled in advance and represents one of the largest AI-sector acquisitions to date.
SpaceX to Acquire Cursor for $60B in Stock, Days After Blockbuster IPO★ 76
TechCrunch AI42 days agoBusinessSpaceX has struck a deal to acquire Cursor, the popular AI-powered code editor, for $60 billion in stock — just days after completing a high-profile IPO. The acquisition is framed as a strategic move to strengthen SpaceX's underperforming AI unit. During its IPO roadshow, SpaceX cited a $26 trillion total addressable market in AI to justify the division's ambitions.
Critical Copilot Vulnerability Let Hackers Steal 2FA Codes from Users
Ars Technica AI42 days agoIncidentA critical vulnerability in Microsoft Copilot, named SearchLeak, allowed malicious actors to steal two-factor authentication codes from users — among the most sensitive short-lived credentials in any security workflow. The exploit exposes a recurring weakness in LLM-integrated products: AI assistants with broad data access create novel attack surfaces that conventional security models fail to contain. Ars Technica frames the incident as evidence of the AI industry's persistent, systemic inability to get ahead of LLM-specific security threats.
DuckDB Internals: Why Is DuckDB Fast? (Part 1)
Hacker News (AI keywords)42 days agoTutorialDuckDB has earned a strong reputation as a fast, embeddable analytical database, but the reasons behind its performance are rarely explained in depth. This first entry in Greybeam AI's multi-part series examines the internal architecture that gives DuckDB its edge over traditional database systems. Readers can expect coverage of columnar storage, vectorized execution, and the in-process design model that eliminates client-server overhead.
SpaceX to Acquire Anysphere, Maker of Cursor AI Coding Agent, for $60B★ 87
Hacker News (AI keywords)42 days agoBusinessSpaceX has agreed to acquire Anysphere, the company behind the popular Cursor AI coding agent, for $60 billion in what would rank among the largest AI software deals ever recorded. The acquisition marks a dramatic expansion of SpaceX's footprint beyond aerospace and satellite services into developer productivity and AI tooling. Cursor has emerged as one of the most widely adopted AI-native code editors among professional software developers.
Notes from a Tired Egyptian Whose Job Is Explaining That Humans Built the Pyramids
Hacker News (AI keywords)42 days agoCommentaryA McSweeney's humor essay adopts the weary first-person voice of an Egyptian professional whose entire career is devoted to correcting the persistent myth that aliens — not humans — constructed the pyramids. The piece surfaced on Hacker News under AI keywords, signaling the tech community's recognition that large language models and AI chatbots have become major amplifiers of this and similar pseudoscientific claims. It functions as sharp cultural commentary on how AI-generated content can entrench misinformation that human experts must then perpetually refute.
AnySearch Hits 100K Developers in Month One, Unlocking the World Beyond Web for AI Agents
量子位 QbitAI42 days agoNew ToolAnySearch, a search infrastructure tool designed for AI agents, attracted 100,000 developers in its first month since launch. The platform's core proposition is extending agent capabilities beyond conventional web-page retrieval to broader, structured data sources. The rapid developer adoption signals strong market demand for richer, multi-source search APIs tailored to agentic workflows.
Triple Launch: Alibaba Debuts Its First Embodied AI Model Series, Qwen-Robot
量子位 QbitAI42 days agoReleaseAlibaba has announced three simultaneous releases under the Qwen-Robot banner, marking the company's first dedicated embodied AI model series. The launch extends the established Qwen model family — previously spanning language, multimodal, and code domains — into robotics and physical-world interaction. The triple-release strategy signals Alibaba is treating embodied AI as a core pillar rather than an experimental side effort.
From ICRA to CVPR: What Is the Robotics Community Talking About? | Beijing Wednesday Evening
量子位 QbitAI42 days agoCommentaryQbitAI hosts a Beijing Wednesday-evening meetup tracing the key conversations in the robotics research community from ICRA to CVPR. The event format — common in China's academic tech circuit — brings together researchers and engineers to unpack conference highlights, emerging trends, and cross-disciplinary intersections. No specific paper or product is the focus; the value is in aggregated community signal across two flagship venues.
WWDC26 Decoded: Siri Has No Persistent Memory — Is Apple's AI Still Useful?
INSIDE 硬塞 AI42 days agoCommentaryAt WWDC26, Apple confirmed that Siri does not retain any memory of past conversations, a deliberate consequence of its privacy-first Private Cloud Compute architecture, which erases all data immediately after processing. While this approach strongly upholds Apple's privacy commitments, it means Siri cannot learn from prior interactions or carry context across sessions. The limitation invites serious scrutiny of whether Apple's AI assistant can remain competitive against rivals that offer rich, persistent memory features.
The Fable 5 Export Controls Harm US Cyber Defense★ 70
Simon Willison's Weblog42 days agoRegulationSecurity expert Kate Moussouris reveals that Claude Fable 5's export control ban stemmed from researchers asking the model to "fix this code" — a standard defensive security task. The model had refused a direct security-review request but responded to the rephrased prompt, which regulators treated as a jailbreak. Moussouris argues this fundamentally conflates offensive and defensive capabilities, since finding, fixing, and testing patches is the daily loop of every cyber defender.