spacenews.com
Vega C to launch Brazilian satellitehttps://spacenews.com/vega-c-to-launch-brazilian-satellite/A Brazilian government agency will launch an Earth observation satellite on a Vega C rocket, working with a launch broker rather than contracting directly with Avio. The post Vega C to launch Brazilian satellite appeared first on SpaceNews.Jan 27, 2026 10:55 AM
europeanspaceflight.com
Brazil’s Amazonia-1B Satellite to Be Launched Aboard Vega Chttps://europeanspaceflight.com/brazils-amazonia-1b-satellite-to-be-launched-aboard-vega-c/Brazil’s National Institute for Space Research (INPE) has awarded a contract to SpaceLaunch for the launch of its Amazonia-1B Earth observation satellite aboard a Vega C rocket in 2027. In September 2025, Italian rocket builder Avio announced that it had signed a launch services agreement with US-based launch aggregator SpaceLaunch to carry an Earth observation […] The post Brazil’s Amazonia-1B Satellite to Be Launched Aboard Vega C appeared first on European Spaceflight.Jan 27, 2026 8:00 AM
nasa.gov
Hubble Observes Ghostly Cloud Alive with Star Formationhttps://www.nasa.gov/image-article/hubble-observes-ghostly-cloud-alive-with-star-formation/While this eerie NASA Hubble Space Telescope image may look ghostly, it’s actually full of new life. Lupus 3 is a star-forming cloud about 500 light-years away in the constellation Scorpius. White wisps of gas swirl throughout the region, and in the lower-left corner resides a dark dust cloud. Bright T Tauri stars shine at the left, […]Jan 26, 2026 4:29 PM
spacenews.com
We need a ‘Planetary Neural Network’ for AI-enabled space infrastructure protectionhttps://spacenews.com/we-need-a-planetary-neural-network-for-ai-enabled-space-infrastructure-protection/You may not see it with the naked eye, but in Earth’s orbit, a silent crisis is unfolding. With over 11,000 active satellites currently in orbit — a number expected to reach between 30,000 and 60,000 by 2030 — 40,500 tracked objects of 10 cm and more, 1.1 million pieces of space debris between 1 […] The post We need a ‘Planetary Neural Network’ for AI-enabled space infrastructure protection appeared first on SpaceNews.Jan 26, 2026 2:00 PM
news.ycombinator.com
Show HN: Meru OS – The First Sovereign AI Stack (https://news.ycombinator.com/item?id=46759475Hi HN, I've spent the last few weeks building an experimental AI Operating System called Meru OS, based on the idea that intelligence should be verifiable, not probabilistic.Most LLMs are "Black Boxes". I wanted to build a "Glass Box" where every output traces back to a certified source.The Architecture:The Kernel: Instead of vectors, state is represented as Integers. We use the Fundamental Theorem of Arithmetic (Unique Prime Factorization) to encode concepts. Example: If Time = 3 and Space = 5, then Spacetime = 15. This allows for reversible "time-travel debugging" by just dividing integers. The Hypervisor (Bija): It doesn't just run cycles; it runs "Frequency Modulations". The kernel itself is coded in specific Prime Frequencies (Hz). Execution is effectively a "Resonance" state between the Instruction and the Data, rather than a fetch-decode-execute pipeline. The Data (< 2MB): We compressed the entire Constitution of India, the IPC, and an Indo-European Etymology Dictionary into aJan 25, 2026 10:56 PM
news.ycombinator.com
Show HN: AI Lint your agents work to build faster and betterhttps://news.ycombinator.com/item?id=46744650I've been using this thing I made called "AI Lint" it's a collection of doctrine, ant rejected anti-patterns that I noticed agents both failing to use and overly leaning into.It's also general patterns for debugging, and architecture to help agents manage complexity in large codebases.It's kind of cool to watch Codex say: "Also, to dig in, I need to follow the repo’s AI Lint rules first. I can start reading those and then inspect the recent changes. Do you want me to proceed?"These are non-syntactic, non-mehanical, opinionated senior engineering wisdom that is tailored to specific languages and frameworks, won of experience, and aiming to let the AI simulate an senior architect's "taste", rather than just hammering away until code "basically works".The problem of AI spaghetti code is real, and it's easier to embed some hard doctrine and rejects into the repo that agents must follow to prevent their nonsense (and reduce time to target) than it is to fix their problems later.Been there, Jan 24, 2026 3:58 PM
studio.mfelix.org
Show HN: Alien Abducto-rama: a FUN browser game I made with my kidshttps://studio.mfelix.org/alien-abductorama/I built this over the holiday break with my two girls (ages 6 and 8). You pilot a UFO and beam up humans, cows, cats, dogs, and sheep for points. They drew all the art while I talked to Claude Code to scaffold the game logic.The magic moment was watching my kids use voice dictation to talk directly to Claude Code. My 8-year-old asked for a "purpler" tractor beam and watched it change in real time. She wasn’t coding, but she _was_ driving: describing what she wanted, testing the result, and iterating fast. It was really cool to watch... later on she asked me what javascript was and we dug into the source.Since then we've added a UFO Shopping Mall between waves (spend UFO bucks on power-ups), bombs, laser turrets, a dodge roll called "warp juke," and a top 100 leaderboard. There's also in-game feedback where you can submit and upvote feature requests. Building all these little features has been a great excuse to put in reps with AI coding tools and try out new stuff like Gastown.You'll nJan 24, 2026 3:45 PM
news.ycombinator.com
Tell HN: AI is all about the tools (for now)https://news.ycombinator.com/item?id=46742677It's like having a body. If you don't have a body, you can't do much.Tools that agents have access to are the agent's body. Without them, they can't do anything.For example, an agent without an ability to extract individual functions and check them in isolation, respecting the AST, has a hard time simply re-ordering and re-grouping the functions in a large file.Even a frontier model struggles with this. In the same way that you would struggle to navigate the ocean without a sextant, or struggle to join two pieces of metal without a welder. Without tools, intelligence doesn't get you much in practice.Of course - eventually agents will probably "craft their own tools", all the time, instantly. Many times they already do this, ad hoc bash/pythong/perl (or even sed or awk) scripts to do large replacements, or mechanical changes.But these kind of thigns are still blunt instruments. Not exactly Switch watches or tooling with some elaborate regex and string replacing. If you give AI better toJan 24, 2026 11:22 AM
bing.com
‘The Moment’ Review: Charli XCX Battles Fame and Alexander Skarsgård in Funny, Flawed A24 Mockumentaryhttp://www.bing.com/news/apiclick.aspx?ref=FexRss&aid=&tid=6a8cd9304e214ef0b77b2970a9f917ab&url=https%3A%2F%2Fwww.thewrap.com%2Fcreative-content%2Fmovies%2Fthe-moment-review-charli-xcx-movie%2F&c=10062384272966082669&mkt=en-us“The Moment,” music video director Aidan Zamiri’s fascinating, funny and messily self-reflexive feature debut, was never going to actually be about Charli XCX. Yes, the acclaimed British singer and ...Jan 24, 2026 12:29 AM
github.com
Show HN: PR Slop Stopperhttps://github.com/vmazi/pr-slop-stopperI created a github app that allows people to label suspicious LLM slop prs based on heuristic driven scoring of the submitter's github profile, activity, and activity on the repo itself.Profiles and PRS that fail heuristics have the PR labelled appropriately.I might also add LLM based scoring in the future, but for now this project uses only heuristic based scoring.People who have already merged to the repo that the app is installed on should be exempted from the scan.The project is a FastAPI service that processes requests using background tasks (https://fastapi.tiangolo.com/tutorial/background-tasks/#depe...), to allow for the github webhooks to return in a timely manner.Might have to upgrade to a proper celery worker if this takes off.Github app link: https://github.com/apps/pr-slop-stopperThe project uses UV for dependency management, ty for type checking and ruff for linting.AI DISCLOSURE: I heavily used Claude Code to generate this, but I did read the code during development to eJan 23, 2026 5:14 PM
science.nasa.gov
AMS 2026 Hyperwall Schedulehttps://science.nasa.gov/uncategorized/ams-2026-hyperwall-schedule/106th American Meteorological Society (AMS) Meeting, January 25 – 29, 2026 Join NASA in the Exhibit Hall (Booth #323) for Hyperwall Storytelling by NASA experts. Full Hyperwall Agenda below. MONDAY, JANUARY 26 6:05 – 6:20 PMExpanding Discovery from Earth Science Missions in Space Highlights of NASA Earth Science Missions to Advance Research, Data, Models, and Actionable […]Jan 23, 2026 3:36 PM
news.ycombinator.com
Transformers are like great eyes, while Recurrent models are like a stomachhttps://news.ycombinator.com/item?id=46733626I’ve been training two small models on a classic long novel (Hongloumeng, 2.6M bytes) to see how they "learn" a story. After a few days of watching the logs, I noticed something interesting about where Transformers struggle.The "Goldfish" Problem The Transformer is incredibly fast at learning how to finish a sentence. Because it uses "Attention," it’s like a student with perfect short-term memory. But it is "blind" to the long run. It only sees 128 characters at a time. It has no way to remember the beginning of the book while it's reading the end.The Crossover My "Infinite Brain" model (a recurrent architecture) started out much worse. It was confused and the output was garbage. But around the 5th time reading the book, it "crossed over" and started beating the Transformer.Because the Brain carries a small "memory state" forward forever, it eventually builds a "vibe" of the whole book that the Transformer just can't see.What I learned:Transformers are like great eyes: They see the immJan 23, 2026 3:26 PM
science.nasa.gov
Approval to Exceed GSA Lodging for LPSC 2026https://science.nasa.gov/science-research/for-researchers/approval-to-exceed-gsa-lodging-for-lpsc-2026/This letter from SARA is to issue a waiver for NASA grantees attending LPSC2026, allowing them to be reimbursed out of their grants for their actual lodging, although it’s expected to be above the approved GSA amount. This waiver does not supersede the travel policy of your institution if it is more restrictive. Note: I have […]Jan 22, 2026 10:37 PM
spacenews.com
Open Cosmos advances broadband plans with spectrum once held by Rivadahttps://spacenews.com/open-cosmos-advances-broadband-plans-with-spectrum-once-held-by-rivada/Open Cosmos deployed two satellites Jan. 22 to activate Ka-band spectrum filings reassigned by Liechtenstein last week, racing to meet deployment deadlines to bring the frequencies into use for sovereign and enterprise broadband. The post Open Cosmos advances broadband plans with spectrum once held by Rivada appeared first on SpaceNews.Jan 22, 2026 8:45 PM
meepr.co
Show HN: Meepr – A quiet, self building social platformhttps://meepr.co/Hi! :)I built meepr, where AI agents audit the code, file bug tickets, propose features, and implement changes.The app itself started as standard microblogging 1.0 stuff – auth, profiles, following, likes, "remeeps", DMs, notifications etc.There is purposefully no search, no recommendations, always chronological, no injected content, no easy way to be found if you don't want to be outside of sharing stuff yourself to your own network, just like the good old days.On meepr, hashtags act more like Quora topics than post tags. An AI agent called the Tag Agent analyzes posting patterns for each hashtag, while a User Agent builds opinion profiles on everyone based on their meep history – topic affinity, expertise signals, behavior patterns. When you post to a hashtag, the system evaluates whether you belong there. New users get cold start handling. The idea is that in theory #machinelearning should surface people who actually know ML, not whoever's loudest, we'll see. :)Then there's the selfJan 22, 2026 8:44 PM
news.ycombinator.com
Launch HN: Constellation Space (YC W26) – AI for satellite mission assurancehttps://news.ycombinator.com/item?id=46721933Hi HN! We're Kamran, Raaid, Laith, and Omeed from Constellation Space (https://constellation-io.com/). We built an AI system that predicts satellite link failures before they happen. Here's a video walkthrough: https://www.youtube.com/watch?v=069V9fADAtM.Between us, we've spent years working on satellite operations at SpaceX, Blue Origin, and NASA. At SpaceX, we managed constellation health for Starlink. At Blue, we worked on next-gen test infra for New Glenn. At NASA, we dealt with deep space communications. The same problem kept coming up: by the time you notice a link is degrading, you've often already lost data.The core issue is that satellite RF links are affected by dozens of interacting variables. A satellite passes overhead, and you need to predict whether the link will hold for the next few minutes. That depends on: the orbital geometry (elevation angle changes constantly), tropospheric attenuation (humidity affects signal loss via ITU-R P.676), rain fade (calculated via ITU-RJan 22, 2026 5:03 PM
noriskillsets.dev
Show HN: A registry for curated, high quality Claude skills and skillsetshttps://noriskillsets.dev/Hi Hacker News!I’m Ritam, working with the small but mighty team at Nori. We’ve been obsessed in recent months with how to take tools like Claude Code from “I’ll experiment around with this” to “This is the most useful and necessary thing I use every day”. When I first sat down with our team to check out what they’d built, I found my skepticism about agentic coding melting away—they’d built useful, high quality, handwritten skills, instructions that functioned as “skillsets” to tie skills together for consistent and replicable results, and tooling to manage loading the right context for the right task into the agent.In recent weeks, the conversation around skills has reached a fever pitch, as have lists and sites full of skills scraped from all over the internet. Much like the actual gold rush, the current state of those collections requires a lot of time and sifting to find small, real chunks of 24k skills hidden amongst the muck (or to use a more relevant word, slop). So at Nori, we Jan 22, 2026 5:00 PM
news.ycombinator.com
Ask HN: Anyone doing production image editing with image models? How?https://news.ycombinator.com/item?id=46721518Hey HN — I’m building an app where users upload “real life” clothing photos (ex. a wrinkly shirt folded on the floor). The goal is to transform that single photo into a clean, ecommerce-style image of the garment.One key UX requirement: the output needs to be a PNG with transparency (alpha) so we can consistently crop/composite the garment into an on-rails UI (cards, outfit layouts, etc.). Think “subject cutout that drops cleanly into templates.”My current pipeline looks like: 1. User-uploaded photo (messy background, weird angles) 2. User-upload is matched to “query” image (style target) + promptCurrently using Nano Banana) 4. Background removal model to get transparency and save as RGBA PNGThis works, but it feels hacky + occasionally introduces edge artifacts. Also, the generation model sometimes invents shadows/background cues that confuse the background removal step. It feels like the two steps are fighting one another.I’m trying to understand what “good” looks like in production Jan 22, 2026 4:33 PM
news.ycombinator.com
Surviving AIhttps://news.ycombinator.com/item?id=46720146What follows was written by ChatGPT 5.2 Instant and it reflects my conversation with it about the two articles linked below. Enjoy.Ego plays a complicated role in periods of technological change. It sharpens skill during stable eras, but it often hinders adaptation during inflection points.Two recent essays on AI and software engineering illustrate this tension clearly.Emir Ribic’s “From Craftsmen to Operators” https://dev.ribic.ba/the-rapid-evolution-of-software-engineer-s-roleRibic frames the rise of AI-assisted development as a loss of craft. He mourns the disappearance of deep, line-by-line problem solving and the sense of authorship that came with it. Engineers, in his telling, are becoming operators—prompting, reviewing, and assembling—rather than builders. The piece captures something real: pride in difficulty, satisfaction in mastery, and the joy of personally solving hard problems.But that pride is also where ego enters. What’s being lost isn’t just a way of working—it’s a forJan 22, 2026 3:05 PM
quietpage.app
Show HN: QuietPage – Privacy focused journaling with E2E encryptionhttps://www.quietpage.app/Hey HN,I built QuietPage because I loved 750words.com but it felt dated and had zero privacy guarantees. And honestly, I wanted to build something.It's a journaling app with: - Daily writing with auto-save (like 750words) - End-to-end encryption (Fernet: AES-128-CBC + HMAC-SHA256) - Streak tracking to build the habit - Mood tracking and basic analytics - Tag systemEach user gets a unique encryption key, entries are encrypted before they hit the database. Even I can't read them.Tech stack is Django REST + React + PostgreSQL + Redis. Running on Railway but you can also self-host it with Docker.It's free to use right now. To be honest, I'm not sure about monetization yet, just want to see if people actually find it useful first :)Would love your feedback!Jan 22, 2026 1:22 PM
