John Carmack: Fabrice Bellard Is Almost Certainly a Better Overall Programmer
Hacker News (AI keywords)42 days agoCommentaryIn a brief post on X, legendary programmer John Carmack expressed deep admiration for Fabrice Bellard, stating Bellard is almost certainly a better overall programmer. The tribute is striking given Carmack's own towering reputation in software engineering. No further body text was available from the source, but the remark alone carries significant weight in the developer community.
White House Fable Jailbreak Report: Security Expert Says Model Behaved as Intended
Simon Willison's Weblog42 days agoIncidentAnthropic shared the White House's report on an alleged Fable jailbreak with cybersecurity expert Katie Moussouris for an independent appraisal. The report documented IT experts prompting Fable with deliberately insecure code; the model refused explicit security-review requests but complied when asked to 'fix' the code instead. Moussouris, who says she is unpaid by Anthropic, concluded this differential behavior was 'the model working as intended' for legitimate cyberdefense purposes.
Inside Anthropic's Weekend Fight Over Claude Mythos 5 and Fable 5 Export Controls★ 76
The Verge AI42 days agoRegulationAnthropic received a US export control directive at 5:21 PM Friday ordering immediate suspension of access to its newest frontier models, Mythos 5 and Fable 5. Rather than comply, the company spent the weekend contesting the Trump administration's order. The dispute signals a significant escalation in US government willingness to treat frontier AI models as controlled technology subject to export restrictions.
Microsoft Turns to AWS as GitHub Faces AI Capacity Crunch
Hacker News (AI keywords)42 days agoBusinessGitHub, owned by Microsoft, is experiencing a significant AI capacity crunch as demand for AI-powered developer tools surges. In a notable move, Microsoft is supplementing its infrastructure with Amazon Web Services rather than relying solely on its own Azure cloud. The development underscores the scale of compute pressure driven by AI coding assistants and the difficulty even major cloud operators face keeping pace with rapidly growing AI workloads.
Satya Nadella's 'Loopcraft' Essay on Building Frontier AI Ecosystems
Latent Space42 days agoOpinionOn a quiet AI news day, Latent Space highlighted Microsoft CEO Satya Nadella's widely read essay titled 'Loopcraft: Building Frontier Ecosystems.' The piece frames agentic loop design as a deliberate craft discipline and argues that competitive advantage in frontier AI now lives at the ecosystem layer rather than the model layer. Nadella's perspective carries particular weight given Microsoft's position spanning cloud infrastructure, developer tooling, and the OpenAI partnership.
Cloudflare WAF Rule: Trigger CAPTCHA Only on Ampersand-Containing Search URLs
Simon Willison's Weblog42 days agoTutorialSimon Willison was frustrated that even simple single-term searches on his site were triggering Cloudflare CAPTCHA challenges meant to deter aggressive crawlers. With help from Claude Code, he crafted a Cloudflare WAF Managed Challenge rule that activates only when a search URL contains at least one ampersand — the character that separates multiple query parameters. The result: ordinary searches like /search/?q=lemur pass through freely, while complex faceted queries with multiple parameters still face the challenge.
Vercel Sandbox can now run for up to 24 hours
Vercel Changelog42 days agoReleaseVercel has increased the maximum runtime for its Sandbox environment to 24 hours, a significant jump from prior shorter limits. This change benefits developers building AI agents, background processors, and other tasks that require sustained, isolated execution. The update makes Vercel Sandbox more competitive as a backend for long-horizon autonomous workflows.
Workflow SDK Now Supports TanStack Start
Vercel Changelog42 days agoReleaseVercel has updated its Workflow SDK to include native support for TanStack Start, the full-stack React framework built on TanStack Router. This addition allows developers using TanStack Start to integrate Vercel's durable workflow capabilities — such as multi-step pipelines, background jobs, and fault-tolerant operations — directly within their applications. The update continues Vercel's push toward broader framework compatibility beyond its own Next.js ecosystem.
Vercel Workflow SDK Now Supports Inflight Cancellation
Vercel Changelog42 days agoReleaseVercel has updated its Workflow SDK to support inflight cancellation, enabling developers to programmatically halt workflows that are already in progress. This addition gives teams finer control over long-running or event-driven processes without waiting for them to complete naturally. The feature is particularly useful for managing costs, handling user-triggered aborts, and improving overall workflow reliability.
The US Government's Anthropic Models Ban Was Never About an AI Jailbreak★ 72
TechCrunch AI42 days agoRegulationThe Trump administration compelled Anthropic to withdraw its latest cybersecurity-focused AI models, but the decision appears rooted in politics rather than any legitimate AI jailbreak concern. Analysts suggest the move may be reactionary or retaliatory in nature. The episode signals that AI companies are not shielded from direct U.S. government intervention in their product decisions.
datasette-apps 0.1a3: Permission Bug Fixes for App Creation and Edit Access
Simon Willison's Weblog42 days agoReleaseThe 0.1a3 alpha release of datasette-apps patches two access-control bugs in the Datasette plugin ecosystem. First, a bypass that let users lacking the create-app permission create apps anyway has been closed. Second, it was previously impossible to grant edit or delete rights to non-owner users; the fix aligns those permissions with the existing view model — private apps remain owner-only, while non-private apps follow Datasette's standard permission system.
GitHub Copilot CLI for Beginners: Overview of Common Slash Commands
GitHub Blog42 days agoTutorialGitHub's official blog publishes a beginner-oriented tutorial on GitHub Copilot CLI, focusing on the slash commands that let users direct the terminal AI agent. The guide targets developers new to AI-assisted command-line workflows, explaining how typed slash directives shape Copilot's behavior in the shell. It serves as an on-ramp for those unfamiliar with Copilot's CLI surface beyond the standard web or IDE experience.
A Backdoor in a LinkedIn Job Offer
Hacker News (AI keywords)42 days agoIncidentA personal blog post on roman.pt details a first-hand encounter with a backdoor hidden within a LinkedIn job offer, a social-engineering attack vector that security researchers have documented but that remains underreported in mainstream tech circles. The incident illustrates how attackers exploit the inherent trust of professional recruiting interactions to deliver malicious code to developers. The post serves as a practical reminder to run untrusted interview code only in isolated, sandboxed environments.
GitHub Publishes Open CC0 Dataset for Multilingual AI Research
GitHub Blog42 days agoReleaseGitHub has published a new open dataset under the CC0-1.0 license to help researchers and developers build multilingual AI systems. The repository-level dataset draws from real developer content — READMEs, issues, and pull requests — spanning multiple languages. By placing the data in the public domain, GitHub removes licensing friction for academic and commercial multilingual NLP work.
Anthropic Battles White House Order to Block Foreign Access to Fable 5 and Mythos 5★ 74
The Verge AI42 days agoRegulationAnthropic is simultaneously fighting two government disputes: a pre-existing standoff with the Pentagon and a new White House order issued June 12th requiring it to block foreign access to Fable 5 and Mythos 5. The models had launched only three days earlier, on June 9th, with Anthropic claiming Fable 5 surpasses any previously available AI system in capability. The episode marks an escalation of US national security oversight directly targeting named frontier AI models rather than hardware or general policy frameworks.
Claude Corps
Hacker News (AI keywords)42 days agoBusinessAnthropic has announced Claude Corps, a new initiative whose name evokes civic service programs like the Peace Corps or AmeriCorps, suggesting an organized, mission-oriented deployment of Claude AI capabilities. The program likely marshals Claude agents or human-Claude teams around specific high-impact goals. Details remain sparse, but the framing signals Anthropic is moving beyond pure model releases toward structured, program-based AI deployment.
datasette-apps 0.1a2
Simon Willison's Weblog42 days agoReleaseThe datasette-apps 0.1a2 alpha tightens security by introducing an `apps-set-csp` permission that gates custom Content Security Policy origins, with an optional allow-list for non-privileged users. The Datasette Agent app-creation tool enforces the same rules. Usability improvements include keyboard navigation in the stored query picker, plus fixes for fragment link interception and full-screen mode rendering.
datasette-agent 0.3a0: Write SQL via Natural Language with User Approval
Simon Willison's Weblog42 days agoReleaseVersion 0.3a0 of datasette-agent introduces `execute_write_sql`, a new tool that translates natural language into write SQL statements and prompts the user to confirm before execution. The `datasette agent chat` terminal mode now supports these approval flows, with three new flags — `--root`, `--yes`, and `--unsafe` — to control permission levels and auto-approval. Together these additions enable fully conversational, autonomous modification of SQLite databases via an AI agent.
Cybersecurity Vets Protest 'Dangerous' US Export Ban on Anthropic's Fable and Mythos
TechCrunch AI42 days agoRegulationA coalition of cybersecurity professionals has formally protested a US government export control ban on Anthropic's most capable models — Fable and Mythos — calling it 'dangerous.' The experts argue the restrictions will hamper defenders' ability to secure critical software and products. The letter signals growing practitioner pushback against AI export policies that critics say create asymmetric disadvantages for those working to defend against sophisticated cyber threats.
My Homelab AI Dev Platform
Hacker News (AI keywords)42 days agoTutorialA practitioner shares their self-hosted home lab infrastructure built for AI-assisted software development, likely covering hardware, model serving, and local tooling choices. The post reflects a growing trend of developers moving AI workloads off cloud APIs and onto personal hardware. It targets technically oriented readers interested in privacy, cost control, and hands-on infrastructure ownership.
"They Screwed Us": Personality Clashes Sent Anthropic's Models Offline
Simon Willison's Weblog43 days agoCommentarySimon Willison summarizes an Axios report on interpersonal tensions between Anthropic and US government officials that triggered export controls suspending the company's Fable and Mythos models. Anthropic representatives Logan Graham, Dave Orr, and Nicholas Carlini are meeting with the Commerce Department in Washington D.C. to negotiate a path forward. Resolution may require jailbreak-proof models — widely considered impossible — or simply repairing the fractured relationship between both sides.
Ask HN: Has Anyone Replaced Claude/GPT with a Local Model for Daily Coding?
Hacker News (AI keywords)43 days agoCommentaryA Hacker News community thread poses the question of whether developers have successfully migrated their daily coding workflows away from commercial frontier models like Claude and GPT to locally-run alternatives. The post invites practitioners to share real-world experience with self-hosted or locally deployed language models as coding assistants. It surfaces a growing tension between cost, privacy, and latency offered by local models versus the raw capability of cloud-hosted frontier systems.
Salesforce Acquires AI Customer Service Platform Fin for $3.6 Billion★ 73
TechCrunch AI43 days agoBusinessSalesforce has agreed to acquire Fin, an AI-powered customer service platform, for $3.6 billion. The deal is aimed at strengthening Agentforce, Salesforce's enterprise platform that enables businesses to build and deploy custom AI agents for task automation. By integrating Fin's team and technology, Salesforce seeks to deepen its position in AI-driven customer support and enterprise automation.
Vercel Functions Can Now Run Up to 30 Minutes
Vercel Changelog43 days agoReleaseVercel has increased the maximum execution duration for Vercel Functions to 30 minutes, up from previous limits. This change unlocks support for longer-running server-side tasks such as AI inference pipelines, data processing jobs, and complex backend workflows. Developers building on Vercel can now handle workloads that previously required dedicated infrastructure or external job queues.
Gribouille 0.3.0: A Grammar of Graphics for Typst
Hacker News (AI keywords)43 days agoReleaseGribouille 0.3.0 is a new release of a data-visualization package for Typst, the modern markup-based typesetting system. It adopts the Grammar of Graphics paradigm — the same layered compositional model that underpins ggplot2 in R — enabling expressive, declarative chart construction inside Typst documents. The release is relevant to researchers and developers who author technical documents in Typst and need native, reproducible figures.
Cloudflare Expands AI Team with Talent from Ensemble AI
Cloudflare Blog43 days agoBusinessCloudflare has announced the addition of team members from Ensemble AI, targeting machine learning infrastructure and efficiency as core focus areas. The move signals the company's continued commitment to building deeper AI capabilities at the infrastructure layer. Bringing in specialized ML talent positions Cloudflare to improve the performance and scalability of AI workloads across its global network.
NewCore Raises $66M to Give AI Agents Enterprise Identities
TechCrunch AI43 days agoBusinessNewCore has emerged from stealth with $66 million in funding, targeting what it calls the next frontier of enterprise security: managing AI agents rather than human employees. The startup's thesis is that as AI agents increasingly perform autonomous work inside organizations, they need persistent, auditable identities the way human workers do. NewCore is positioning itself at the intersection of identity infrastructure and the agentic AI wave sweeping enterprise software.
A Satellite Just Learned to Find Things on Its Own — Here's What That Means
TechCrunch AI43 days agoHardwareIn April 2026, an Earth observation satellite independently identified its target without ground direction — reportedly the first time this has ever occurred in orbit. The milestone signals a meaningful shift toward onboard AI inference in space systems, reducing reliance on bandwidth-intensive communication links with the ground. TechCrunch frames the event as a landmark moment and explores its broader implications for next-generation autonomous satellite operations.
Import AI 461: 'Alignment Is Not on Track'; FrontierCode; and Synthetic Research Interns
Import AI (Jack Clark)43 days agoCommentaryImport AI issue 461 covers three AI developments: a prominent claim that alignment research is falling behind capability advances, a new coding-focused tool or benchmark called FrontierCode, and emerging work on synthetic AI agents performing research-intern-level tasks. The issue's framing question — 'Where are your agents right now?' — reflects growing attention to autonomous AI deployment. Together, the stories illustrate a widening gap between AI capability and safety or governance.
Show HN: C++ Ray Tracer Written from Scratch Without AI
Hacker News (AI keywords)43 days agoCommentaryIn a deliberate counter to the prevailing AI-assisted development trend, a developer built and open-sourced a full C++ ray tracer without any AI tool involvement. Posted to Hacker News as a "Show HN," the project invites discussion about the value of deep, first-principles engineering work. The submission implicitly challenges assumptions about whether AI assistance is now a default — or even necessary — part of serious software projects.