Read Tech News With Digest
Helps you quickly pick interesting articles to read.
Cloudflare's Clef decision models beat Jev on accuracy and latency, now open sourceCloudflare released Clef and Clef-flash, two decision models that produce typed probabilistic outputs for agentic workflows. Clef leads the Jev Decision Index and is 2x faster than gpt-oss-120b in a threat intelligence test. The models are open-sourced under Apache 2.0 on Hugging Face and hosted on Workers AI. Cloudflare also launched an RL fine-tuning service to adapt Clef to specific use cases.
"In music theory, a clef is a symbol placed at the beginning of a musical staff that assigns specific pitch names to the lines and spaces. A decision model is analogous to a music clef because it helps define the domain of the context and the subsequent notes (actions) that follow it."
Turbopuffer abandons vector-primary storage to end write amplificationTurbopuffer is rebuilding its storage engine, moving from a vector-primary layout to one where ANN search becomes just another secondary index. The old design caused storage and write amplification and capped block sizes, limiting non-vector query performance. The new engine, turbopuffer v3, passes all CI tests but currently regresses performance; the team is now tuning it and sharing benchmark results as they go.
"The solution to these problems is simple: don't key on the ANN address. That is precisely the change turbopuffer v3 makes."
After years of development, StreetComplete — the OpenStreetMap survey app — is now in public beta on iOS. The port uses Kotlin Multiplatform and Compose Multiplatform to share code between Android and iOS, with the migration about 50% complete. The project relies on community contributions and funding to reach a full release.
"We very roughly estimated some time ago that developing a full port to iOS would take one man-year of work. So, the scope of this remains very large."
Bez is a project that generates a web rendering engine from specifications, using the three major browsers and WPT as tests. The goal is to make building a browser engine cheap enough that many can exist, not just a few. Early results show generated CSS layout rules pass all 227 test cases, and three-browser voting can identify real compatibility bugs like Firefox's 1/60px rounding.
"Building a web engine by hand costs hundreds of engineers and many years, so only a few companies can have one, and only they decide how the web works."
Ask HN: Who is hiring? (October 2026)I'm posting the monthly 'Who is hiring?' thread for October 2026. Please state the location and include REMOTE for remote work, REMOTE (US) or similar if the country is restricted, and ONSITE when remote work is not an option. Only post if you personally are part of the hiring company—no recruiting firms or job boards—and if you are actively filling a position and committed to replying to applicants.
"Snout | Multiple Engineering + Product Roles | Remote US or Ontario, Canada | Full Time. Join us at Snout on our mission to ensure no one ever has to make a health decision for their pet based on the cash in their bank account. Snout plans pay for 100% of routine veterinary care, unlimited visits, and additional member benefits - think pet insurance, that you will actually use every year. We're growing the team and hiring across the board. Featured role this month: Founding Data Engineer - be our first data hire: make sense of hundreds of millions of rows from vet practice-management systems and our own app, and turn our Postgres and dbt warehouse into a foundation the whole company trusts. Our tech stack includes Node.js, React, PostgreSQL, AWS, Tailwind, and Python. We use both JavaScript and TypeScript heavily. Our team uses LLM-based tooling day to day, but we're fundamentally a human-centric team, and our product itself does not heavily feature AI components. Apply at https://jobs.gem.com/snout"
"Open Education Applications / Neon | Senior/Lead Platform & DevOps Engineer, Senior Frontend Engineer, Senior Full-Stack Engineer | Utrecht, The Netherlands | HYBRID | DUTCH REQUIRED. Please note: working proficiency in Dutch is strictly required for these roles, and we kindly ask for no automated applications. The rest of this post will continue in Dutch. Wij zijn een non-profit die zich inzet om lesmateriaal voor scholen beter, betaalbaarder en flexibeler te maken, in zowel digitale als gedrukte vorm. Werktijden en kantoordagen zijn flexibel en worden in overleg bepaald op basis van de rol. Wij bieden een competitief salaris. Technologiestack: Git monorepo, TypeScript, Yjs en React, een beetje Python, OpenTofu/Terraform, Scaleway, managed PostgreSQL en managed Kubernetes. Openstaande rollen: - Senior/Lead Platform & DevOps Engineer - Senior Backend Engineer/Architect - Senior Frontend Engineer - Senior Full-Stack Engineer. E-mail jobs[at]openeducation.foundation met (HN) + de functienaam in het onderwerp, samen met een korte introductie en een link naar je LinkedIn-profiel en/of GitHub. Als er van beide kanten een goede match lijkt te zijn, plan ik graag een kennismakingsgesprek in. https://openeducation.foundation | https://www.neon.nl"
"We The Flywheel | Production Designer | REMOTE (async, global) | Contract, part-time to full-time | https://wetheflywheel.com/en/careers/production-designer-remote/ We're a small, agent-led studio. Most of our ops run through AI agents, and the people here direct them. We need a production designer who turns briefs into polished, on-brand marketing, social and ad creative, fast, inside an existing brand system. What we're looking for: - Real Figma experience (required) - You've wired Figma into coding agents like Codex or Claude Code (e.g. Figma MCP), so a design becomes working code without a hand-off - Strong typography and layout craft, and you can keep quality up across a high volume of assets - Clear written English for async work. Apply through the link. It starts with a short skills assessment, no cover letter needed. For other roles, see our careers page: https://wetheflywheel.com/en/careers/"
"PAGNOS | Cloud Engineer, AWS Serverless (TypeScript/Node.js) | REMOTE (Germany) | Full-time | €75k–80k | business-fluent German required. We build serverless applications on AWS for clients and for our own SaaS products. On client projects, we work directly with their developers and help them understand and maintain what we build. You’ll write backend code in TypeScript/Node.js, build APIs with Lambda and API Gateway, model data in DynamoDB, and design event-driven workflows and integrations. Infrastructure lives in code using AWS CDK, with CI/CD pipelines for builds and deployments. You’ll help decide how systems are designed and take testing, observability and security into account as you build. There’s some React/Next.js work too. We’re looking for practical experience building AWS serverless applications with TypeScript/Node.js, including infrastructure as code, CI/CD and event-driven design. The role includes working through requirements and technical decisions with clients, so you should be comfortable with that. We use coding agents daily and expect you to use them too, and to share useful approaches with other developers. AWS serverless development is the main focus of the role. We’re remote-first, use Scrum, and work with code reviews, pair programming, and unit and integration tests. We meet in person regularly, sometimes in Cologne, sometimes in Austria or Mallorca. We plan travel around family commitments. You must be based in Germany and speak business-fluent German and good English. We don’t offer visa sponsorship. €75k–80k, 32 days of vacation, hardware of your choice, and company-paid AI tool subscriptions. Send your CV to recruiting@pagnos.com – no cover letter needed. Questions before applying are welcome. Full job ad (German): https://germantechjobs.de/jobs/PAGNOS-GmbH-Cloud-AI-Engineer--AWS-Serverless-mwd"
"PrairieLearn (Remote US) — Full-Stack Software Engineer — TypeScript / Postgres / React / AI. PrairieLearn (https://www.prairielearn.com) is an open-source assessment platform used by universities across the US (Berkeley, Princeton, Michigan, Illinois, and others). We power mastery-based learning and large-scale exams. We’re a small, profitable, early-stage company (bootstrapped, no VC) that is growing quickly. Our users love us and we have very high retention and rapid spread through word of mouth. As an early-stage hire, you’ll work across the stack and enjoy meaningful ownership from day one. Tech we use: Node.js / TypeScript backend, Postgres, AWS, React. We’re also developing AI tooling, including LLM agents to help instructors create content, and vision-language models to help grade student work. PrairieLearn is open core: https://github.com/PrairieLearn/PrairieLearn Details: - Location: Remote (US only) - Salary: $100k-$180k depending on experience - Benefits: Stock options (0.5% - 1.5%), unlimited PTO, flexible hours - Role: Full-time. (We’re not able to sponsor visas at this time.) Apply at https://www.prairielearn.com/jobs-ashby"
"Shepherd (Series B) | ONSITE | San Francisco, CA & New York City, NY. At Shepherd, we're pursuing the most ambitious technical vision in commercial insurance: fully autonomous underwriting. Shepherd is an AI-native commercial insurance platform transforming how high-hazard industries get covered. The infrastructure behind the AI boom (data centers, semiconductor fabs, renewable energy assets) has to be built and insured, but traditional carriers weren't built for this speed. We built Shepherd to solve that. We announced our $42M Series B earlier this year! That brings our total funding to over $60M — led by Intact Private Capital, the investment arm of one of the largest insurers in the world. Intact is not only our lead investor but also a carrier partner, a testament to the confidence the incumbent industry has in what we're building. We’re looking for product-minded, high ownership engineers to join our team! The Engineering team is HQ’d in SF, but we’re growing in NYC now too! We’re hiring for 2 Senior Full Stack Software Engineers, 1 Machine Learning Engineer and 1 AI Product Manager in SF, and 2 Staff Full Stack Software Engineers in NYC. Engineers own projects end-to-end so full stack ability is expected. Check out our open roles here: https://shepherdinsurance.com/careers Our stack includes: Typescript, React, Next.js, GraphQL w/ Apollo, Node.js, Postgres & Redis – If you have experience building AI agents or multi-step reasoning systems, we’d love to hear from you. * In the form, be sure to mention that you heard about this opportunity via Hacker News. If you want to read more about our team and culture, check out our blog: https://shepherdinsurance.com/blog"
RacketCon 2026, the sixteenth gathering for the Racket programming language, takes place this Saturday and Sunday with both in-person and streaming options. Highlights include a keynote by Pat Hanrahan, Turing Award winner and RenderMan designer, plus talks on type inference, low-level programming with Pille, IDE-friendly macros, floating-point accuracy with Herbie, effect handlers, OOP with BrandX, immutable dataframes, and a new FFI. A town hall closes the event.
"Floating-point math requires rounding, so it’s not perfectly accurate. But how inaccurate it is depends on how you do your computation: two ways of writing the same mathematical formula can have radically different accuracy."
Weave Router - Open-source model routing for coding agentsWeave Router is an open-source model router that plugs into any coding agent, like Claude Code or Codex, and intelligently switches between LLMs to match the task. It uses an ensemble approach to deliver frontier-level performance at a fraction of the cost. Weave Router 2.0 matches GPT-6 Astra on Terminal Bench 4.0 and SWE Atlas, while cutting costs by nearly half and running tasks over twice as fast. Built with a hidden Markov model and classifier for smart routing, it's available on GitHub or as a hosted service.
"Matching frontier model performance was a huge milestone for our routing model, and in my opinion validates our initial hypothesis that an ensemble of models can do better than any single model ever could."
Cloudflare launches K2, a serverless event stream built on R2 object storageCloudflare has launched K2 in public beta, a durable event streaming service built on R2 object storage. It decouples producers and consumers, scales horizontally, and supports long-term retention. Under the hood, K2 implements a partitioned log on R2, achieving 11 9s durability while simplifying operations. The beta is available for Workers Paid accounts with 10GB storage and 30 MB/s produce limits. Pricing will be $0.04/GB produced and consumed, and $0.02/GB/month retained.
"We first built K2 because we needed a durable buffer on the edge, initially to serve as the ingestion layer for Basin Pipelines."
ESP32 Chips Hide a Secret SDR Mode That Turns Them Into 80 MS/s Spectrum AnalyzersThe ESPARGOS team discovered an undocumented feature in several ESP32 chips that bypasses WiFi and Bluetooth to capture raw IQ baseband samples, turning the microcontrollers into internal SDRs covering 2.2–2.7 GHz (plus 4.8–6.0 GHz on the ESP32-C5) with up to 80 MS/s and 13–54 MHz of analog bandwidth. Two other projects independently found similar capabilities, though continuous streaming requires an FPGA or the new ESP32-S31.
"The team also says phase-coherent transmissions would be possible, but they aren't implementing them right now because they could be misused."
Ask HN: Who wants to be hired? (October 2026)I am posting the monthly 'Who wants to be hired?' thread for October 2026. I ask job seekers to share their location, remote preference, relocation willingness, technologies, resume/CV, and email. I request that only individuals looking for work post, and that readers only email to discuss opportunities.
"Location: Malaysia / nomading in Southeast Asia Remote: Yes, experienced with US/EU hours and async teams Willing to relocate: Within Southeast Asia Technologies: Python, SQL, PostHog, GTM, GA4, Shopify Web Pixels, n8n, analytics/attribution, technical SEO, landing pages Résumé/CV: https://lightyoruichi.com/work?password=octhn182 Email: harinder.mke@gmail.com LinkedIn: https://www.linkedin.com/in/lightyoruichi"
"I'm a full-stack marketer and developer. For about 15 years, I've worked where customer insight, growth strategy, and implementation meet. Companies usually bring me in when the product is good but growth is blocked. Positioning may be unclear, the funnel may be leaking, attribution may be broken, or niche's where there's no existing playbooks. I go through the customer data, sales calls, support messages, product usage, and analytics. Then I find the real problem and build what is needed: the positioning, landing pages, tracking, automations, content, or first growth test."
"Some things I've done: - Grew a professional-services company from zero to 210k monthly organic visits and 20 to 222 qualified leads a month. - Interviewed 20+ potential customers for an AI company, found two use cases, and helped get two enterprise design partners before launch. - Built SEO, analytics, tracking, landing pages, and acquisition systems for SaaS, ecommerce, and service companies. - Managed performance marketing across Southeast Asia with budgets up to $2.5m a month."
"Looking for contract, fractional, or full-time work in growth, product marketing, GTM, growth engineering, or marketing operations."
"Location: Porto, Portugal Remote: Yes; occasional office visits Willing to relocate: No Technologies: Node.js, TypeScript, GraphQL, REST, PostgreSQL, MongoDB, ClickHouse, Kafka, Redis, AWS, Docker, Kubernetes, Terraform, OpenTelemetry, Vue.js, React Résumé/CV: https://thaler.dev/thaler.cv.pdf Email: dmitry@thaler.dev"
"Senior full-stack engineer with a backend focus and 15+ years of experience. Most of my work is at the integration layer: bringing awkward external systems together behind GraphQL and REST APIs that product teams can rely on. This has included payment gateways and government services at a national lottery, multi-protocol network telemetry, and retail planning systems processing billions of sales and inventory records. I take ownership beyond the application code, including infrastructure, deployment, and observability. I care about keeping systems reliable and cloud costs under control. EU-based B2B contractor; no sponsorship required. Experienced working remotely with US companies via Deel and Globalization Partners. Available with 7 days’ notice."
"Location: Hanford, CA (Pacific Time, US Citizen) Remote: Yes (Preferred) Willing to relocate: Yes (SF Bay Area, San Diego, PNW, or West Coast tech hubs) Technologies: Python (FastAPI, Django, Pytest), Go, Ansible, Rootless Podman, systemd Quadlets, Traefik, GCP (Cloud Functions, Secret Manager), GitOps, Linux Résumé/CV: https://resume.delpilar.net GitHub: https://github.com/redjordan2539 LinkedIn: https://linkedin.com/in/jordan-del-pilar/ Email: jordan@delpilar.net"
"Backend & Platform Automation Engineer with 4+ years of experience building API middleware, GitOps infrastructure, and event-driven automation. I bridge the gap between application software and systems administration—writing clean Python, securing endpoints, and automating manual workflows."
"Selected Repos & Impact: - Chiron-Relay (github.com/redjordan2539/Chiron-Relay): Async serverless FastAPI middleware on GCP Cloud Functions bridging Twilio webhooks and Gmail API. Features raw payload HMAC-SHA1 request validation, GCP Secret Manager integration, and Pytest coverage in CI. - johto-infra (github.com/redjordan2539/johto-infra): Declarative Ansible IaC repository deploying rootless Podman Quadlets bound natively to systemd unit files. Features Traefik TLS cert resolution, Ansible Vault secret management, Gitea Actions CI/CD with ansible-lint, and automated Restic DR pipelines with explicit retention policies. - Production Dispatch Engine: Built a Python automation tool to parse daily delivery workflows and send customer SMS alerts—cutting a multi-hour manual task down to under 30 minutes."
"US Citizen based in CA with zero visa requirements. Looking for mid-level Backend, Platform, or SRE / Infrastructure Automation roles (Remote, or West Coast Hybrid/On-site)."
"Location: Paris / Fontainebleau, France Remote: Yes, EU/CET (hybrid Paris ok) Willing to relocate: No Technologies: TypeScript, React, Node.js, LLM agents, MCP servers, Claude Code, Codex, Figma Plugin API, TouchDesigner, WebGL/GLSL Résumé/CV: https://www.somaticbits.com/David_Pettersson_Senior_Creative... Email: hn@somaticbits.com"
"Creative technologist / product engineer. Two sides: AI tooling and automation for teams, and generative/immersive art. Same instinct either way - prototype fast, make the abstract real."
"Recent work: - Idea Engine (open source): start from a seed word, wander an association map, and pin the path you took into a buildable concept card. Three-model pipeline - one proposes associations, one scores and filters them for surprise and coherence, one writes the card. Svelte + Hono + SQLite, runs locally in Docker. https://github.com/somaticbits/idea-engine - Forward-deployed engineer at a large hospitality/marketplace company. Built automations that removed up to 6h of manual work per project. - Slides and websites generated straight from existing Figma components. Design system in, finished artifact out. - Currently writing a custom MCP server for a calisthenics training app, so workouts and progress are coached and queried by an agent rather than through the UI."
"Looking for full-time work where AI tooling meets creative tech. Teams building something new rather than iterating on what's there. AI & automation: https://www.davidpettersson.com Generative & immersive art: https://www.somaticbits.com"
"Location: Spain Remote: Yes Willing to relocate: Prefer remote, open to travel and occasional onsite Technologies: Scala, Java, Kotlin, Go, Python, Ruby, Rust, TypeScript/React, Postgres, AWS, GCP Résumé: https://alexitc.com/resume LLM-readable résumé: https://alexitc.com/llms.txt Email: alexis+hn@alexitc.com"
"10+ years, generalist. I pick the tool for the problem instead of forcing one stack everywhere. Most of my work is backends, data and event pipelines, full-stack product work, crypto, and rescuing broken or legacy systems. I've led teams of up to 12. I use Claude Code daily on real production work, generate-and-review, and build my own AI tooling. If you want a fast read on whether I fit, paste my llms.txt into an LLM with your job description and ask where I match and where I don't. My site has an "Ask AI" button that does the same. Available immediately and used to working on US timezones. US C-Corp for contracting."
"Location: Europe (Netherlands, CET) Remote: yes Willing to relocate: no Technologies: Go, Python, Kotlin, PostgreSQL, Redis, Kubernetes, Terraform, GCP, AWS Résumé/CV: https://robinsiep.com/experience Email: contact [at] robinsiep.com"
"Senior backend engineer focused on distributed systems and system design, with about a decade of end-to-end ownership across design, implementation, and production. Most recently, I was lead backend engineer and architect on a distributed platform for venture capital workflows (Go services, event-driven architecture on GCP, data reconciliation, observability), and lead engineer on a survey fulfilment system integrating various third-party providers. Both roles included mentoring and coordinating other engineers. I'm looking for a permanent role, but remain available for contract work in the meantime."
Researchers introduce Context Language Models (CLMs), which treat context as a file the model can update freely, learning what to keep. Built zero-shot with existing models, CLMs outperform state-of-the-art context management: 11.4% higher accuracy with 21.5% fewer FLOPs on BrowseComp-Plus, 5% higher scores with 59% fewer FLOPs on 12-hour EdgeBench, and 65% greater improvement on a 24-hour multi-repository agent-swarm task. Natural-language instructions and online RL further boost performance, while Suffix Cache Reuse cuts serving compute by 35%.
"We introduce Context Language Models (CLMs), language models that natively manage their own context."
Rust compiler performance improved by 4.57% in mean wall-time between July and September 2026, with 555 of 629 benchmarks showing gains. Key contributors include PGO for Clippy, LLVM 23, and optimizations to the new borrow checker and trait solver. The author also highlights a dataflow analysis change that cut compile time by 30% on a large crate. He will join Hexcat to work on compiler performance.
"The mean wall-time reduction was 4.57%, which is a remarkable improvement in just two months. Of the 629 benchmark measurements, 555 of them improved and only 74 regressed."
Papero extracts PDF structure with plain geometry, no ML modelsPapero is an open-source PDF parser that recovers reading order, tables, formulas, figures, and bounding boxes using only CPU and geometric algorithms—no ML models or PyTorch. It runs in the browser, as a Python package, or as a REST API, and processes dense arXiv papers at 39 ms per page with zero failures on 54 papers. Output includes Markdown, JSON, Word, and Excel with full layout fidelity.
"Getting the text out of a PDF is easy. Getting its structure back — which column comes first, which lines are a table, where the formula is — is what makes the output usable for RAG, search and LLMs."
Polyedergarten: A Garden of Paper Polyhedron ModelsPolyedergarten is a website by U. Mikloweit that showcases an extensive collection of paper polyhedron models, including Platonic and Archimedean solids. The site, active since 1992, offers downloadable templates and VRML models across 28 pages, inviting enthusiasts to explore and build these geometric shapes.
"In my garden you find polyhedron models made of paper: Platonic polyhedra, Archimedean polyhedra and other polyhedron models."
OpenAI and Synopsys Team Up to Build GPT-Synopsys, an AI That Designs ChipsOpenAI and Synopsys have signed a multi-year partnership to develop GPT-Synopsys, a specialized model that learns to operate Synopsys' EDA tools like an expert engineer. It will run on OpenAI infrastructure and integrate with Synopsys.ai and Synopsys Autopilot, letting engineers delegate design objectives and iterate toward verified outcomes. The deal includes revenue sharing, joint go-to-market, and enterprise-grade data protection.
"We're using our most advanced technology to improve the systems that power AI. With Synopsys, we're bringing that work to chip design, helping engineers explore more designs and get to a working chip faster."
Gmail Is Bouncing Your App's Email? Here's How to Fix SPF, DKIM, and DMARCIf your automated emails land in spam or bounce with a 550 5.7.26 error, your sending domain probably lacks proper authentication. This guide walks through setting up SPF, DKIM, and DMARC DNS records, explaining how each check reads a different domain and why alignment with your visible From address matters. It includes concrete examples for a subdomain like mail.example.com, with verification steps and common pitfalls like duplicate SPF records.
"This explains the otherwise confusing result where SPF and DKIM both say “pass,” but DMARC says “fail.” The checks can succeed for the wrong domains."
ParadeDB Closes the Gap with PlanetScale's TIN, Then Pulls AheadAfter PlanetScale launched TIN, a fast full-text search extension for Postgres, ParadeDB responded with two key optimizations: storing fieldnorms alongside postings to eliminate random access, and implementing a MAXSCORE pruning algorithm for multi-term disjunctions. On a 28.7M Hacker News dataset, ParadeDB now outperforms TIN by 2x on top-k BM25 queries. The post also examines benchmark configuration differences, including TIN's dense-term elision, which trades correctness for speed.
"TIN is fast (at least 8x faster than ParadeDB 0.25 in every PlanetScale benchmark). So fast that the only response which made sense was to shut up and put on our performance optimization hats."
The OpenID Foundation's new whitepaper argues that today's OAuth 2.1 and OpenID Connect frameworks work for simple, single-domain AI agents but fail for cross-domain, asynchronous, or highly autonomous ones. It identifies urgent gaps: agent identity fragmentation, user impersonation, consent fatigue, recursive delegation, and browser agents bypassing API controls. The report calls for interoperable identity profiles, explicit on-behalf-of delegation, and new programmatic verification for trustworthy autonomy.
"Agents controlling visual interfaces directly (or via MCP into browser orchestrators) bypass all traditional API-based authorization controls. Protecting the open web from lockdown will require robust authentication of web bots or web agents."
OpenDLSS reimplements Nvidia's DLSS 5 neural rendering network in Vulkan, bit-exactA Vulkan reimplementation of Nvidia's DLSS 5 neural rendering network, matching the original bit-for-bit. The same 71-block Swin/ViT network runs FP8 on tensor cores, with all 75 block boundaries byte-identical. A WebGPU port achieves the same exactness in a browser without tensor cores or FP8. You supply the weights. Performance reaches 2.8 ms at 768x768 on an RTX 4070 SUPER.
"The exactness is in the specification, not in the hardware."
IBM Is Dissolving Red Hat, One Badge at a TimeTechrights reports that IBM is phasing out Red Hat's brand and culture, with employees leaving en masse. A former Red Hat staffer notes his badge now says IBM. The article argues that redhat.com blog posts sell slop rather than Linux, inflating the share price without substance, and that IBM is creatively imploding.
"The "Red Hat" brand (and badge) is being dissolved some more today."
Micron CEO: Memory Supply Will Be Much Tighter in 2027 and 2028 Than in 2026Micron CEO Sanjay Mehrotra warned on the company's fiscal Q4 2026 earnings call that memory and storage supply will be significantly tighter in 2027 and 2028 than in 2026. With record quarterly revenue of $54.2 billion and over 75% of 2027 output already committed, Micron sees no return to supply-demand balance. DRAM prices jumped in the high teens last quarter, NAND about 30%, and new fab capacity won't arrive until late 2028.
"Even with additional industry DRAM clean room space plans with robust demand trends, including new upside requests from customers, we do not have line of sight to when supply and demand will return to balance."
GrayKey Freezes iPhones to Defeat Apple's Anti-Forensics RebootMagnet Forensics, maker of the GrayKey phone-unlocking tool, has developed GrayKey Preserve and Evidence Preservation Mode to bypass the iPhone's inactivity reboot, which normally locks out police after 72 hours. Leaked training video shows the tool forces Airplane Mode to cut off radios and keep the phone in an After First Unlock state, preserving sensitive data indefinitely. A security researcher calls it a game changer, and the ball is now in Apple's court.
"That AFU state is captured. Even if that device does reboot for any number of reasons, memory maintenance or the power is lost or whatever, the AFU state is not lost. This is the true magic behind the GrayKey Preserve and the Evidence Preservation Mode function."
Figma's remote MCP server only accepts clients on its supported list, and Pi isn't on it. When David Soria Parra, who created MCP, asked Figma to fix the restriction, Figma's Gayani replied that Pi could be considered if a form is filled out. Parra called the restriction sad, saying he envisioned MCP as an open ecosystem. Others compared it to letting only Firefox access a website.
"I think in spirit, when I created MCP, I envisioned an open ecosystem. That to me feels core. Seeing restrictions like this is sad and I hope Figma can get to a point where it's more open or at least make the process of getting into the allowlist very easy."
Book of Shapes: A Collection of Minimal, Generative, and Customizable SVG PatternsBook of Shapes is a curated collection of minimal, generative, and customizable SVG patterns created by Nikolaj Sokolowski. It features over 100 patterns across categories like grid, radial, noise, flow, isometric, organic, distortion, and physics. Each pattern is designed to be easily customized and used in various projects.
"A collection of minimal, generative and customizable SVG-patterns."
Inside SlutCon: A Lawyer's Weekend as a Flirt Girl at a $3,300 Seduction RetreatOverlocked, a 26-year-old lawyer, recounts three days at SlutCon, a Berkeley retreat where men paid up to $18,000 to practice seduction on paid 'flirt girls.' She describes rationalist subculture, panels on canthal tilt and rejection reframing, and a tense moment when a male attendee told a flirt girl, 'If I were a shooter, I'd have a girlfriend.' Despite aggressive feedback, she never felt unsafe.
"You are trying to get a million dollars. The million dollars is stored in his balls."
Gemini 4 Argon writes 1M tokens in one go, cracks a healthcare flaw other models missedGoogle's Gemini 4 Argon is a frontier model for long-horizon reasoning, rolling out first to trusted cyber defenders. It sets state-of-the-art scores on DeepSWE v1.1 (77.9%), LVBench (91.7%), and CWE-bench v1 (68%), and can generate up to 1M output tokens. Internally, it optimized quantum subroutines by 40%, freed 300 TiB of memory, and migrated C/C++ code to Rust. Pricing starts at $2 per million input tokens.
"In an early demonstration of its impact, the model uncovered a critical vulnerability exposing sensitive personal information across healthcare software used by hospitals worldwide, identifying a severe risk that previous frontier models had missed."