AI 资讯

Overview Tulisan ini menceritakan pengalaman saya meng-install NixOS di VPS Jetorbit memakai nixos-anywhere , dideploy dari laptop macOS (nix-darwin). Komposisi Laptop (macOS + nix-darwin) : Semua konfigurasi NixOS dibuat di sini, lalu dikirim ke server. VPS Jetorbit (x86_64 Linux) : target deploy. Boot lewat legacy BIOS , dengan IP statis (bukan DHCP). nixos-anywhere : dipakai sekali saja untuk install awal. nixos-rebuild : dipakai untuk update konfigurasi setelah install awal. Catatan Penting : VPS tidak perlu menyimpan file konfigurasi sendiri. Kita edit config di laptop, lalu push ke VPS lewat SSH. Tidak ada apt update , tidak ada config drift, dan kondisi server selalu reproducible dari git. Pre-flight: Cek Sebelum Menyentuh Server Banyak kegagalan install sebenarnya bisa dicegah kalau kita cek tiga hal ini: 1. Mode boot: UEFI atau legacy BIOS? Ini menentukan konfigurasi bootloader. Salah konfigurasi di sini dapat menyebabkan GRUB gagal saan install nanti. Di VPS (SSH ke VPS): [ -d /sys/firmware/efi ] && echo "UEFI" || echo "BIOS" Di Jetorbit secara default adalah BIOS . Artinya konfigurasi GRUB mode legacy BIOS, bukan EFI. 2. Networking: IP, netmask, dan gateway Jetorbit memakai IP statis, bukan DHCP. Kalau konfigurasi NixOS kita set IP maka NixOS akan fallback ke DHCP, dan akan gagal dapat IP, sehingga server jadi tidak bisa diakses sama sekali setelah reboot. Di VPS: ip -4 a # lihat IP + CIDR dan nama interface (misal. ens3) ip route # lihat "default via X.X.X.X", itu gateway-nya Contoh: 2: ens3: ... inet 110.XXX.XXX.XXX/24 brd 110.XXX.XXX.255 scope global ens3 3. SSH Root nixos-anywhere memerlukan akses root ke server untuk bisa melakukan instalasi. Akses ini didapat lewat SSH key, bukan password. Jadi pastikan Di VPS: public key sudah terdaftar di root . Konfigurasi SSH root ini bisa dilakukan saat install OS awal di panel Jetorbit atau dikonfigurasi secara manual di ~/.ssh/authorized_keys . Cara Kerja nixos-anywhere Sebelum menjalankan perintahnya, ada ba

AI moderated interviews that read how people feel Discussion | Link

This is the fourth post in Craft & Code , a short Friday series about what carpentry can teach us about AI, skill and the future of software. Last week I worried about where the next generation's judgement will come from. This week, why we may not notice it is missing until it is too late. My father built me shelves in an alcove when I was small, and I mentioned in the first post that they may still be there for eternity. The other side of that story is the one every household knows: the shelf that is not quite right. The one that sags under a row of books, or sits a degree off true so that anything round rolls gently to one end. You do not need to be a carpenter to see it. A bad joint, a door that will not close, a shelf that dips — the material tells on the maker, immediately and to everyone. That is the comforting version of the analogy, and the one I expected to write: carpentry is honest about its failures because they are visible, while software can look polished and be rotten underneath. A wonky shelf looks wonky; bad software looks finished. It is a tidy line, and there is real truth in it. But it is only half the truth, and the more interesting half should worry us — because the moment you go up from a shelf to a serious piece of engineering, the comfort falls away completely. Consider two of the most admired structures of the last century. The Tacoma Narrows Bridge was designed by one of the leading suspension-bridge engineers of his day: elegant, slender, celebrated. It opened in the summer of 1940 and tore itself apart in the wind that November, twisting like a ribbon because the design had not reckoned with how the deck would behave aerodynamically. Nobody had seen a wonky bridge; it looked magnificent. The flaw was real, fundamental, and invisible until the wind found it. The Citicorp Center in New York, finished in 1977, was a triumph of structural engineering, raised dramatically on great columns at the midpoints of its sides. Only after it was compl

Where AI incidents in legal actually come from, and what infrastructure (not policy) prevents them. Blake Aber · Predicate Ventures · 2026 The policy layer is table stakes. It isn't enough. When Sullivan & Cromwell apologized to a federal bankruptcy judge in April 2026 for AI hallucinations in a court filing, the firm's apology letter said the firm had policies. Safeguards existed. Those safeguards weren't followed. That framing, "the safeguard existed but wasn't followed," is how a policy failure gets described. But something more specific happened: a hallucination was generated, wasn't caught at generation time, wasn't caught at review time, and made it into a document that got filed. That's not a policy problem. It's an infrastructure problem. The distinction matters because it determines what you build next. What policy can and can't do Policy is a promise made before the event. A well-written AI acceptable-use policy says: don't submit output you haven't reviewed; verify citations before they go into a document; a human must approve anything client-facing. This works when the human executing the task has time, attention, and professional accountability in that moment. It fails when one of those is missing: a deadline, a junior practitioner, a late-night run. Policy can't: Verify a citation at the point of generation Flag output that has drifted below a confidence threshold Stop hallucinated text from appearing in a draft before a human ever sees it Detect when the underlying model is behaving differently than it was in testing Policy can: Set the expectation that review must happen Define who bears accountability when it doesn't Create a paper trail after the fact One of those is prevention. The other is compliance. What infrastructure does instead An AI harness layer operates at the point of generation, not at the point of review. This reflects a broader reality that production AI is mostly harness and very little model . For legal work specifically, three com

Semantic HTML and Accessibility: Building Better Websites Introduction Semantic HTML is the practice of using HTML elements that clearly describe the purpose of the content on a webpage. Instead of using many <div> elements, semantic tags such as <header> , <nav> , <main> , <section> , <article> , and <footer> make the page easier to understand. Semantic HTML is important because it improves accessibility, helps search engines understand web pages, and makes code easier to read and maintain. Before: Non-Semantic HTML <div class= "header" > <h1> My Portfolio </h1> </div> <div class= "navigation" > <a href= "index.html" > Home </a> <a href= "about.html" > About </a> </div> <div class= "content" > <p> Welcome to my portfolio website. </p> </div> After: Semantic HTML <header> <h1> My Portfolio </h1> </header> <nav> <a href= "index.html" > Home </a> <a href= "about.html" > About </a> </nav> <main> <section> <p> Welcome to my portfolio website. </p> </section> </main> Accessibility Issues I Found 1. Images Missing Alternative Text Before: <img src= "images/profile.jpg" > After: <img src= "images/profile.jpg" alt= "Profile picture of Grace Loko" > Adding alternative text allows screen readers to describe images to users with visual impairments. 2. Navigation Was Not Semantic Before: <div> <a href= "index.html" > Home </a> <a href= "about.html" > About </a> </div> After: <nav> <a href= "index.html" > Home </a> <a href= "about.html" > About </a> </nav> Using the <nav> element helps assistive technologies identify the website navigation. 3. Form Inputs Had No Labels Before: <input type= "text" placeholder= "Your Name" > After: <label for= "name" > Name </label> <input type= "text" id= "name" name= "name" > Labels improve accessibility by helping screen readers identify each form field. Conclusion This accessibility audit helped me understand the importance of semantic HTML and accessible web design. By replacing non-semantic elements with semantic tags, adding image alt text,

Why AI programs at PE portfolio companies stall at the same organizational seam, and what to do about it. Blake Aber · Predicate Ventures · 2026 There's a failure mode I've watched play out at enough portfolio companies that I've given it a name: the ownership dyad. It goes like this. The AI program is running. The product manager owns the roadmap (what the AI should do). Engineering owns the deployment (how it does it). Both parties are competent. Both are aligned on the goal. And the AI initiative quietly stalls anyway, usually somewhere between the promising pilot and the production system that was supposed to follow. The mechanism is diffuse accountability at the decision layer. What the dyad looks like in practice In the average portco planning meeting, the PM and the engineering lead sit across from each other. The PM has a change request: "The model is producing summaries that miss the key clause in contracts above a certain length. We should fix this." Engineering hears this and wants to know: is this a prompt change or a model change? Either requires scoping, and scoping requires the PM's input on acceptable behavior. So engineering asks the PM. The PM says "whatever's best technically." Engineering ships a prompt change. The next month, the same issue appears in a different context. The PM brings it back. Neither person is wrong. Neither person is slacking. The problem is structural: there's no single person who can describe (precisely and completely) what the AI should produce, evaluate whether it's producing it correctly, and approve a change to the system without requiring the other party's sign-off. The dyad looks like shared ownership. It functions as diffuse accountability. No one is in charge of the model's behavior. The failure mode at month nine Most portco AI programs that make it through a successful pilot still die quietly around month nine of production. The most common reason is not that the model got worse. It's that the harness around the m

Galfus Script has reached its first MVP milestone. Galfus is an experimental programming language written in Rust, designed around a typed VM-first execution model, compact .gfb artifacts, deterministic module/workspace resolution, and an ownership model based on anchors, edges, and weak observers. The MVP goal was not to build a full ecosystem yet. The goal was to prove the complete local execution pipeline: txt .gfs source -> lexer and parser -> resolver -> type checker and semantic analyzer -> ownership check -> MIR lowering -> bytecode emitter -> Galfus Module Image -> .gfb serialization -> VM interpreter execution https://github.com/vulppi-dev/galfus-script/discussions/10

[v0.3.0] - 2026-06-29 🚀 Added Checkpoint Mechanism — ReActCore introduces three checkpoints during streaming: content_ended (after text content), before_tool_calls (before tool calls), and after_tool_calls (after tool calls), enabling precise interception and state synchronization of the execution flow. Message Queue System — Added a new MessageQueue class in run.py , supporting async enqueue, drain, and remove operations. Users can now queue messages while the LLM is running; queued messages are sent automatically after the current task completes. The frontend introduces a QueueBar component to display queued messages, with CSS spinning animation, single-line ellipsis, and hover-to-delete functionality. Queue Message Merging — MessageQueue.drain_all() now merges consecutive messages with the same name into a single message, preventing fragmented user input when multiple queue entries share the same sender. Queue WebSocket Events — The execution event protocol introduces three new event types: message_queued , queue_drained , and queue_returned ( useRunWebSocket.ts ). The frontend processes queue state updates in real time. Stop & Queue Integration — When the user clicks Stop, pending queued messages are returned to the input box via queue_returned . Checkpoint stops cleanly clear the queue and automatically start the next message. System Notification Messages — Introduced the SystemMessage type (with notification role) to separate error messages from assistant content. Errors are now rendered as independent notification bubbles, no longer embedded within assistant message cards. tiktoken Real-Time Token Estimation — ReActCore initializes a tiktoken encoder on startup for real-time token counting during streaming. Unknown models fall back to o200k_base . 🔧 Improved Custom Model Name Auto-Complete — The model name field in ModelManager has been upgraded from Select to AutoComplete , allowing users to type custom model names not in the predefined list. Message Block T

Qwen's new image paper is easy to read as another benchmark bump. Qwen-Image-2.0-RL takes the existing Qwen-Image-2.0 model, runs a reinforcement-learning pass on top, and reports better scores: 57.84 on Qwen-Image-Bench, up 2.61 points from the base model. Its text-to-image arena Elo moves from 1115 to 1193. Its image-editing arena Elo moves from 1256 to 1349. Those are the headline numbers. They are not the useful part. The useful part is the training story underneath them. The paper is a good reminder that "just optimize the reward" is a dangerously incomplete sentence, especially when the model is not an LLM and the output space is a whole image. The model got better, but not by one simple trick Qwen-Image-2.0-RL is a post-training pipeline for a diffusion image model. In plain English: the base model already knows how to generate and edit images. The RL stage tries to steer it toward outputs humans prefer, including better prompt following, better aesthetics, better portrait fidelity, and more reliable editing. The team builds task-specific reward models. For text-to-image, those rewards cover alignment, aesthetics, and portrait quality. For editing, they cover instruction following and face identity preservation. Then they train with a GRPO-style setup adapted for flow-matching diffusion models. If you only squint at that, it sounds like the same broad recipe people use for language models: generate candidates, score them, push the model toward the better ones. The paper is more interesting because it shows how fragile that story becomes once you touch the actual training loop. The CFG detail is the first real lesson Classifier-free guidance, usually shortened to CFG, is one of those diffusion-model knobs that users mostly experience as "make the image follow the prompt harder." Under the hood, it changes how the model samples. The Qwen team tested three ways to use it during RL. Using CFG during both rollout and training made the images collapse into incohere

I built this project as an entry for the H0: Hack the Zero Stack with Vercel v0 and AWS Databases Hackathon. #H0Hackathon Inspiration Justice moves slowly. I learned that firsthand as my family navigated a legal dispute. What struck me wasn't just the stress — it was that things were quite disorganised. Documents were paper-based or buried somewhere in emails. Updates came through WhatsApp messages. Simple documents took a really long time to draft and send. The system was fragmented and difficult to navigate. Companies like Harvey tackle document drafting well, but legal research tools and LLM wrappers can hallucinate case law, citing judgments that don't exist. I knew that if I was going to build something for this space, it had to be grounded in real, verifiable law. That led me to Laws Africa, which provides structured access to actual South African legislation and court judgments. I also noticed a problem that lawyers experience daily: the mechanical work. Logging into court portals to file a case. Hunting through OneDrive, Google Drive, and Dropbox for the right version of a document. Sifting through hundreds of emails to find something relevant to a matter. Onboarding a new client when the intake form is a PDF someone emails you. These are not AI problems. They are automation problems — and lawyers or their secretaries are doing them manually every single day. That became Agently. What Agently Does Agently is a legal workspace that handles the full lifecycle of a matter, from the moment a client submits an intake form to the day the case closes. Matter Management. Every client engagement lives in a structured matter. Documents, emails, notes, contacts, workflows, and AI conversations are all scoped to it. A lawyer can open a matter and immediately see everything relevant. AI Agent with Real Legal Research. The AI connects to Laws Africa's knowledge bases — South African legislation, court judgments, and municipal law — so research is grounded in actual legal

Want to go back to your ex... AI assistant? We don't blame you. Take these steps to get Google Assistant back after your fling with Gemini.

The Supreme Court just placed once-independent agencies more firmly under presidential control. The court ruled in Slaughter v. Trump with a 6-3 vote that President Donald Trump had the authority to fire the Federal Trade Commission's two Democratic commissioners, even though it broke with decades of prior legal precedent at the time. The justices have […]

Comcast plans to spin out NBCUniversal and leave the entertainment business.

The all-stock deal values Iridium at $8 billion, and gives Rocket Lab even more firepower to compete against Amazon and SpaceX.

Asymm Systems has released Eliya 25.0.3, an OpenJDK 25 LTS distribution aimed at improving production diagnostics in Java environments. It consolidates several HotSpot features into an opt-in Production profile. Eliya is designed for teams needing reliable diagnostic data, especially in regulated settings. Future enhancements are planned for Phase 2. By A N M Bazlur Rahman

Enterprise investment in AI is booming. Gartner is calling 2026 an “inflection year” for organizations to align their AI projects with strategic business objectives. As the pressure to prove ROI mounts, executives and technology leaders are looking to agentic AI to drive the measurable financial outcomes their businesses seek. A prime opportunity for AI agents…