AI 资讯

This isn't really a typical Ask HN, but since my LinkedIn network doesn't get me there, I thought I'd give HN a shot! Paddle, the merchant of record, used to have quite excellent support (widget in their dashboard) which then turned into workable support by e-mail. Now they have added a layer of AI automation whereby any support e-mail receives an automated AI reply to which you have to reply to escalate to a human. Unfortunately that loop is broken. I never get a human into the loop at all (the

From musical instruments and games to play pieces that encourage balance and dexterity, these toys will engage your child’s mind and body.

Hi HN, Gabriel here. This is a standalone binaural beat generator, combined with a two-deck local audio player built with the Web Audio API. It includes ten harmonic carrier pairs, seven brainwave-state presets, per-harmonic mixing (so the binaural tones aren't out-of-tune with the music), adjustable beat and carrier frequencies, two oscillator shapes, offline key/BPM analysis, automatic carrier matching, adaptive impulse noise, playlist support, and output recording. The main problem I've had w

“I think it should stand for artificial idiot,” one 9-year-old says. Here’s why kids of all ages are calling AI “disgusting” and “creepy.”

Building an operating system is a project I have had my eyes set on ever since I discovered free will in the realm of programming. Years ago, I did a reasonable amount of research, paying extra attention to the subject during my computer science degree and I was able to understand Operating System Theory and how it works from first principles but I never really got around to building one. I had only flimsy reasons for not embarking on it like "why build one when there are tons of working ones out there? The theoretical knowledge is enough" . More recently, I am ignoring the need to not re-invent the wheel for the joy of programming. So if you are interested in also rebuilding stuff because you can, join me on this series as I document how I am going to be building kluster. kluster is in its infancy and the direction is not clear but the one certain thing is that I will be building it entirely in Rust, save some assembly instructions and a linker script and I will be explaining every single line of code along the way. It will also be designed to target the raspberrypi 4 & 5, on qemu and on real hardware respectively. This is an opportunity for anyone who wants to see how Rust works at the lowest of levels to hop on and join the ride. Note that this series will be your biggest lesson on delayed gratification because we will write a lot of code before we even get to see anything meaningful on screen but I will foreshadow what you can get by the end of part 3 if you are patient enough: {{ image(src="/images/os-part3-result.png", alt="Part 3 Results OS Dev") }} You can also clone the source code for part 1 from Github and follow along. Project Setup First things first, let us setup the foundation of the project. I'll be straight with you, I love Rust and I enjoy using the Rust ecosystem in its entirety so I will stay true to that and use it as obsessively as any true Rustacean; I won't hold back. Without doubt, all the dependencies we need are freely available as long as

I typed a 19-character question into Claude Code: who are you? Then I opened the network logs of what my machine actually sent over the wire. I recommend you never do this. It ruins something in you. The numbers My question was 19 bytes . The request that carried it was 118,693 bytes . That's a 6,200x amplification before the model reads a single word I wrote. My actual message was 0.016% of the payload. If this were a letter, I'd be mailing a phone book with a Post-it note inside. Here's the pie, byte for byte: What Bytes Share 38 tool JSON schemas 78,935 66.8% System prompt (behavior manual, safety rules, my git status) 22,178 18.8% A list of 60 more tools it could load, plus 25 skill descriptions 16,844 14.3% My memory files, my email address, today's date 1,126 1.0% My question 19 0.016% And the tool schemas aren't even schemas, mostly. Of those 78,935 bytes, only 29,164 are machine-readable JSON schema. The other 44,392 bytes are English prose — instruction manuals for the model, shipped on every single request. One tool alone, Workflow , carries a 19,012-byte description. That's a small essay. It rides along on every message I send, forever, whether I orchestrate a multi-agent workflow or just say "hi." Sixteen of the 38 tools are for driving a browser preview I never opened. There's an iOS Simulator controller in there. I was asking a chatbot its name. The full manifest, since you asked Every tool schema in that request, largest first. "Prose" is the English description; "schema" is the actual machine-readable part. Approximate tokens = bytes ÷ 4. Tool Bytes Prose Schema ~Tokens Did I need it? Workflow (multi-agent orchestration) 21,609 19,012 1,833 ~5,400 No Artifact (publish web pages) 10,679 6,358 4,002 ~2,670 No AskUserQuestion 4,371 1,097 3,149 ~1,090 No iOS Simulator control 3,976 1,307 2,518 ~990 No ScheduleWakeup (self-pacing loops) 3,942 2,685 1,068 ~985 No visualize show_widget 3,038 648 2,259 ~760 No Agent (spawn subagents) 3,037 1,574 1,342 ~760 N

You know the feeling. You call your AI bot in Slack, hit enter, and then... nothing. Dead silence. You sit there staring at the screen. Is it actually running your code? Did the server crash? Is it just quietly judging your prompt? You have absolutely no idea. It's a black box of anxiety. The standard solution to this UX nightmare isn't much better. Most bots just spam the thread with a million separate status messages: "Claude is thinking..." "Claude is running a command..." "Claude is scratching its nose..." Congratulations, your Slack thread now looks like a chaotic WhatsApp group. We got fed up with both extremes. The "Aha!" Moment: Slack as an IDE While working on the claude-slack-bridge , we realized we didn't just want a chatbot. We wanted a local terminal experience, right inside the workspace. So, we updated the bridge with a feature that basically turns Slack into a full-blown IDE. Instead of polluting the channel with endless new messages, everything updates live on a single, dynamic message. Here is what it looks like in action: How the UX actually feels now: Real-time transparency: You literally see it run the commands and update the progress sequentially. No more guessing if it's dead. Zero thread pollution: It edits the same message block over and over. Direct delivery: When it's done, it just spits the final output (like test.pencil.png in our demo) straight into the chat. No noise. No ghosting. Just a clean, live-updating terminal. Try it out If you're building AI integrations or just want to see how we handled the Slack API rate limits to make this live-updating feel smooth, check out the code. 🔗 GitHub Repo: claude-slack-bridge I'm curious – how are you guys handling long-running AI tasks in your Slack apps? Are you Team "Spam the Thread" or Team "Silent Black Box"? Let me know in the comments.

I just wanted to open a CSV file quickly. Instead, I got: Slow spreadsheet apps Online tools that upload my data Way too much friction So I defined a simple goal: Fast and frictionless Fully local (no uploads) Spreadsheet-like experience That’s what I built. What you can do with it Open CSV files instantly (drag & drop, no setup) Search and filter large datasets in seconds Edit data like a spreadsheet Export exactly what you see Why this matters Everything happens locally in your browser: Your data never leaves your device No account required No tracking, no storage Built with client-side JavaScript — no backend involved. Just open, edit, and export. How it works Built with client-side JavaScript — no backend involved. Spreadsheet-Like Editing Click a cell to select it, then press Enter or double-click to start editing. Press Enter again to save the value and move to the cell below. Use Option + Enter on macOS or Alt + Enter on Windows to insert a line break. You can also drag the small fill handle to copy a value across multiple cells. Export the Current View The exported CSV reflects the current table state, including: Search results Sorting order Visible columns Saved cell edits Try it yourself Open a CSV file, edit a few cells, and export it — all without uploading anything. 👉 https://csv-open.github.io/ No signup. No upload. Just works. Handles CSV files up to 25MB Supports multiple languages

As part of my learning journey with Django, I wanted to build a project that would challenge me beyond the basics. I decided to create RecipeHub, a web application where users can create, manage, and share recipes while exploring recipes from other users. The project started from a Django starter template, but I customized it by adding new features, redesigning the interface, and deploying it online. Features RecipeHub allows users to: Register and log in Create, edit, and delete recipes Browse recipes by category Save favourite recipes Upload recipe images Access a personal dashboard Use the application in both light and dark mode The application is fully responsive, making it easy to use on both desktop and mobile devices. Technologies Used I built the project using: Python Django Django Allauth PostgreSQL Tailwind CSS DaisyUI HTMX Vite Gunicorn Render GitHub was used for version control throughout the project. Challenges One of the biggest challenges was deployment. While everything worked locally, deploying to Render required configuring PostgreSQL, environment variables, and static files correctly. I also encountered an issue with uploaded recipe images. Since the application is hosted on Render's free tier, uploaded media is stored on an ephemeral filesystem, meaning uploaded images are lost after redeployment. Learning why this happens gave me a better understanding of the difference between development and production environments. Another challenge was redesigning the dashboards. I wanted them to feel clean and modern instead of looking like a default Django application, so I spent time improving the layout, spacing, and responsiveness. What I Learned This project helped me improve my understanding of: Django project structure Authentication and user management CRUD operations Database relationships Responsive UI design Git and GitHub workflows Deploying Django applications Debugging real-world issues More importantly, it taught me how to troubleshoot proble

If you’re tired of sponsored and AI-summarized search results, Kagi gives you a curated—and human-centric—way to find what you’re looking for.

The headline going around is that Spark 4.2 can retire your vector database. That's half true, which is the most dangerous kind of true. Spark did add real vector search, and for some workloads it genuinely removes a whole system from your stack. For others, you'd regret dropping your vector DB. Here's the honest version. The short answer If your vectors already live in your data platform and your searches are batch or analytical, Spark 4.2 can absolutely replace a separate vector database. If you're serving live, low-latency retrieval for a chatbot or search box, you probably still want a dedicated one. It's a "depends on the workload" answer, and the details matter. What Spark 4.2 actually added Spark 4.2 brought vector search into plain SQL. No bolt-on library, no separate engine. The new primitives include vector distance and similarity functions, vector normalization, vector aggregation like sum and average, and the headline one, NEAREST BY, a top-K ranking join that finds the closest matches by distance. In practice that means you can store embeddings in a Spark table and run a similarity search with SQL you already know. Those operations cover the real use cases: retrieval, recommendations, entity resolution, and candidate generation. Databricks is openly framing this as Spark becoming an AI serving layer, not just a batch engine. Why this is a big deal The value isn't that Spark invented vector search. Plenty of tools do it. The value is that you can keep your retrieval pipeline on one platform. Think about the normal setup today. Your data sits in a lakehouse or warehouse. To do vector search, you spin up a separate vector database, then build a pipeline to copy and sync embeddings into it, and keep the two in step forever. That's a second system to run, secure, pay for, and debug at 2 a.m. Spark 4.2 lets you skip that for a lot of cases. The embeddings stay where your data already is, and the search runs right there. Fewer moving parts is a real win, and i

xAI has a bold pitch for Grok 4.5: it codes about as well as Claude Opus 4.8, but does it with roughly a quarter of the tokens. That's not a "we're smarter" claim. It's a "we're just as good for far less money" claim, which in 2026 might matter more. Someone actually put it to the test, so let me walk through what the numbers say and why you should care. The claim and the pricing Grok 4.5 landed on July 8, 2026. xAI says it matches Opus 4.8 on coding while using about 4.2 times fewer output tokens to get there. The sticker price already favors Grok. It runs $2 per million input tokens and $6 per million output. Opus sits at $5 and $25. That's less than half the price on both sides before you even factor in the token efficiency. Stack the two together and the cost gap gets dramatic. What the benchmarks say The benchmarks mostly support the marketing, with a catch. On Terminal-Bench 2.1, which measures real command-line work, Grok 4.5 scored 83.3 percent to Opus 4.8's 78.9. But on SWE-Bench Pro, the harder test of fixing real open-source bugs, Opus still comes out ahead. So the honest read is not "Grok is better." It's "Grok is about as good, for a lot less." Different claim, and a more interesting one. The hands-on test Benchmarks are one thing, real work is another. The New Stack ran a head-to-head, giving both models the same three jobs in one real Rust project (the fd file-finder) inside Cursor, and tracked every token. I'm summarizing their results here, credit to them for actually measuring it. The three tasks were a bug fix, a multi-file refactor, and a feature build. The code both models produced was nearly interchangeable, so the story came down to tokens, time, and cost. On the small bug fix, Opus actually won. Both wrote an identical fix with all tests passing, but Opus did it faster and on fewer tokens. Grok's efficiency edge showed up on the bigger jobs. On the refactor, Grok used about 197K tokens versus Opus's 954K for the same result, roughly a fifth.

The US government just accused a Chinese AI company of copying an American model, and threatened sanctions over it. Whether or not the claim holds up, it's worth understanding what's actually being alleged, because the word at the center of it, distillation, is about to come up a lot. Here's the plain version. What distillation means Model distillation is training a new model using the outputs of an existing one. Instead of learning purely from raw data, the new model learns by copying how a stronger "teacher" model responds. Done legitimately, it's a normal technique for making smaller, cheaper models. Done against someone else's commercial model without permission, it's a way to clone a lot of that model's behavior on the cheap, and it usually violates the terms of service. That's the core of the accusation: that one company used another company's model as an unpaid teacher. What the White House claims Michael Kratsios, who leads the White House Office of Science and Technology Policy, accused Moonshot AI, a Beijing-based company, of distilling Anthropic's Fable model to build its Kimi K3. He alleged Moonshot built a sophisticated internal platform to run large-scale distillation against US models while switching between access methods to avoid detection. The claim didn't stop there. Kratsios also said Moonshot obtained restricted Blackwell-generation Nvidia servers through Thailand, which would sidestep US export controls. Treasury Secretary Scott Bessent followed up by putting sanctions and export-control blacklisting on the table. Why some experts are skeptical Here's the part that keeps this from being open-and-shut. Anthropic's Fable model has only been public since July 1, and Kimi K3 shipped on July 16. That's a very short window to distill a 2.8-trillion-parameter model primarily from another model's outputs. Several experts have pointed out that the timeline makes "mostly built by distilling Fable" hard to believe. So the honest framing is that this is a

Moonshot launched Kimi K3, and within 48 hours it had to stop taking new subscribers. Not because the model flopped, but because too many people wanted it. That's a strange kind of problem to have, and it's telling you something important about where AI's real bottleneck now sits. What actually happened Less than two days after Kimi K3 went live, Moonshot froze new subscriptions. The reason was blunt: demand had eaten through its available GPU capacity. In the company's own words, the model got "far more love than we expected," and in 48 hours usage pushed close to the limit of what its hardware could serve. Moonshot handled it reasonably. Existing users kept their access, and the company said it would expand capacity and reopen signups in batches. It also split its plans into two tiers, a general Kimi Membership for web and app use, and a separate Kimi Code Membership aimed at programming work. That split is a hint about which users are burning the most compute. Why the model drew that kind of demand Kimi K3 isn't a minor release. It's an open-weight model at 2.8 trillion parameters, and it reportedly beat Anthropic's Fable 5 and OpenAI's GPT-5.6 Sol on front-end coding tests. Open, cheap, and competitive on real coding is exactly the combination developers pile onto. So they did. The real story: the bottleneck moved Here's the part worth internalizing. For years the hard, expensive problem in AI was training. That's where the giant compute bills and the headlines were. Kimi K3's freeze shows the constraint shifting to inference, the cost of actually running the model for users, every request, every day. Agentic workloads are why. When an AI agent runs a coding task for minutes or hours instead of answering a single prompt, each user consumes far more compute than a chatbot ever did. Multiply that by a viral launch and you hit a GPU wall fast. Moonshot didn't run out of ideas. It ran out of chips to serve the ideas. Why this matters even if you never touch Kimi Thi

Microsoft has intervened to stop Windows 11 users with LG monitors from being bombarded with annoying McAfee trial pop-ups. In response to complaints about the LG bloatware, Microsoft's Windows chief, Pavan Davuluri, said that LG has agreed to immediately disable the McAfee pop-up from its LG Monitor App Installer, and pledged that Microsoft will "keep […]