AI 资讯

Over the past few months, I've been building a gaming website focused on Elden Ring guides, calculators, and tools. While the project started as a simple hobby, it quickly became an interesting experiment in SEO, content strategy, and web development. Here are some lessons I learned along the way. Building the Site Was Easier Than Getting Traffic Launching a website with Next.js was straightforward. Getting visitors was much harder. Many developers underestimate how competitive search traffic can be, especially in gaming niches where large sites already dominate search results. Publishing a website is only the first step. Why I Chose Next.js The project uses: Next.js TypeScript React Tailwind CSS The biggest advantage was SEO. Server-side rendering and static generation helped ensure that search engines could easily crawl and index pages. Performance was also excellent compared to many traditional CMS solutions. Tools Attract Different Users Than Articles One interesting discovery was that calculators and interactive tools behave differently from standard content pages. For example: Guides answer questions. Tools solve problems. A player may read a guide once, but they might return to a calculator dozens of times while planning different character builds. This makes tools valuable long-term traffic assets. Internal Linking Matters More Than Expected When new content was published, internal links helped search engines discover and understand related pages. For example: Build guides linked to calculators. Calculator pages linked to stat guides. Stat guides linked to weapon builds. This created a stronger topical structure around the Elden Ring ecosystem. Search Traffic Takes Time One of the biggest lessons was patience. Many pages received: Zero impressions Zero clicks No rankings for days or even weeks. Then suddenly search impressions started increasing as Google tested pages across different queries. Traffic growth was rarely linear. Content Clusters Work Well Inst

Your 24/7 personal AI agent Discussion | Link

OpenAI has rolled out an upgrade for the free model you interact with the most on ChatGPT.

I build Voxis, an open-source Windows app that translates whatever your system is playing — a video, a game, the other side of a call — and plays the translation back as spoken voice, a few seconds behind the speaker. No subtitles, no virtual audio cable, no bot joining your meeting. The "no virtual cable" part is the bit worth writing about. Almost every system-audio tool on Windows tells you to install VB-CABLE or VoiceMeeter, or to drop a bot into your call. Voxis doesn't, for incoming audio. This post is how that capture engine works, and the sharp edges I hit building it in Python. I'll be specific about what's hard and honest about what's not mine to fix. The goal Read the exact audio the user is hearing — the post-mix system output — at 16 kHz mono, and do it without installing anything. Then stream it to a translation model and play the result back, all while the original keeps playing underneath. Three constraints fall out of that: Driverless. If it needs a reboot and a driver, it's not zero-setup. No self-feedback. The app plays translated audio into the same system mix it's capturing . Naively, it would capture its own voice and translate the translation. That has to be impossible by construction, not patched with an echo gate. Realtime-safe. Capture can't stall. If the downstream VAD or garbage collector hiccups, the WASAPI ring buffer must not overflow. WASAPI process-loopback: capturing the mix, minus yourself Windows 10 version 2004 added the ApplicationLoopback API — a way to activate an IAudioClient in loopback mode scoped to a process tree, either including only that tree or excluding it. Excluding our own process tree is exactly what constraint #2 needs: the captured mix is everything the user hears, with Voxis's own output removed. You don't get this client from the normal IMMDeviceEnumerator path. You activate it by name through ActivateAudioInterfaceAsync , passing the loopback parameters in a PROPVARIANT carrying a BLOB : params = AUDIOCLIENT_

And by viral I mean from $0 to $31. Umami told me Clew Directive got 14 visits last month. AWS told me I owed $31 for it. That works out to $2.21 a visitor, which would make it the most expensive free learning-path tool in California. Spoiler alert: 14 visitors, $31, and not a single one of them was the reason. Something was off. Here is how Amazon Q, Claude, and a few hours of reading my own code untangled it. The app turned out to be innocent. What Clew Directive is, quickly A free, stateless tool that builds you a personalized AI learning-path PDF. You take a 60-second Vibe Check, four questions about your goals and how you learn, and it maps you to free, verified resources and hands you a briefing. No accounts, no database, no paywall, nothing stored about you. It runs on Amazon Nova, which is why it costs close to nothing to operate, which is also why a $31 bill made no sense. The name is the Theseus kind of clew. A ball of thread to find your way out of the maze. Less hype, more direction. Live at clewdirective.com . The number that didn't add up Twelve visitors, 14 visits, 93% bounce, average session about a minute. Referrers from Bing, Google, Yahoo, GitHub. Visitors from the US, India, Netherlands, Egypt, Ethiopia, Singapore. Mostly crawlers stopping by to say hello. A few curious humans and a parade of bots is not a $31 month. So either every visit was doing something enormous, or the bill was never about visits at all. The dashboard lied, politely. An Amazon Q Story My cost tracker said Clew Directive was running on Claude Sonnet. Sonnet is the expensive one. Case closed, right? I opened the repo. Clew Directive does not run Sonnet. The Navigator agent runs Amazon Nova 2 Lite. Scout and Curator run Nova Micro. The IAM policy is scoped to Nova ARNs only, so a Sonnet call from these functions would come back AccessDenied. The app physically cannot bill Sonnet. The math agreed. A full learning-path generation on Nova costs about two-tenths of a cent. Fourtee

Live demo: https://safestate.vercel.app , code: https://github.com/usv240/safestate A product recall today is basically a notice. It lives on a webpage, or a PDF, or an email that somebody is supposed to read. Say the problem out loud and it gets uncomfortable fast. A recalled crib can be listed and sold to another family, and nobody in that sale ever sees the recall. Reselling recalled goods is actually illegal, and recalled infant products have killed kids. I spent this hackathon building something to close that gap. I called it SafeState, and the idea is small: make the recall do something. When a second-hand item is listed or sold, the marketplace checks SafeState first, and recalled units get blocked right at checkout. It is precise down to the serial number, so safe units still sell. It runs on the stack this hackathon is about. A Next.js front end on Vercel, with Amazon Aurora DSQL behind it. Why DSQL is the whole point here The promise SafeState has to keep is this: the moment a recall lands in any region, no marketplace anywhere should ever read that product as "safe" again. That is a strong consistency problem, not a nice-to-have. If there is any window where a recalled product still looks safe, that is exactly when it gets sold. An eventually consistent store or a nightly sync leaves that window open. DSQL's active-active, multi-region setup with strong consistency is what closes it. I set up a real peered cluster across us-east-1 and us-east-2, with us-west-2 as the witness. Write a recall through one region's endpoint and you can read it back from the other region right away. There is a page in the app that lets you run that yourself. The one trick that makes it work DSQL runs on snapshot isolation (PostgreSQL REPEATABLE READ) with optimistic concurrency. It catches write-write conflicts at commit time. Snapshot isolation will not protect you from write skew, so I had to design around that. To guarantee that a recall and a sale of the same product actua

What just happened On June 22, 2026, AWS quietly launched Lambda MicroVMs. Not a Lambda feature update. A new compute primitive sitting between Lambda Functions (stateless, 15-min max) and EC2 (full VM, you manage everything). Each MicroVM is an isolated Firecracker VM with its own HTTPS endpoint, running your code from a pre-built snapshot. Stateful. Up to 8 hours. Suspend when idle, resume on demand. I tested it the same week. Here's what I found. The test setup A minimal Python HTTP server packaged as a Dockerfile: from http.server import HTTPServer , BaseHTTPRequestHandler import json , time , os class Handler ( BaseHTTPRequestHandler ): start_time = time . time () request_count = 0 def do_GET ( self ): Handler . request_count += 1 body = json . dumps ({ " message " : " Hello from Lambda MicroVM! " , " uptime_seconds " : round ( time . time () - Handler . start_time , 2 ), " requests_served " : Handler . request_count , " pid " : os . getpid () }) self . send_response ( 200 ) self . send_header ( " Content-Type " , " application/json " ) self . end_headers () self . wfile . write ( body . encode ()) HTTPServer (( " 0.0.0.0 " , 8080 ), Handler ). serve_forever () The Dockerfile: FROM public.ecr.aws/lambda/microvms:al2023-minimal RUN dnf install -y python3 && dnf clean all WORKDIR /app COPY app.py . EXPOSE 8080 CMD ["python3", "app.py"] How it works Three steps: Zip code + Dockerfile → upload to S3 create-microvm-image builds the container, starts the app, takes a Firecracker snapshot of memory and disk run-microvm launches from that snapshot Every launch resumes from the pre-initialized state. No cold boot. Your app is already running the moment the MicroVM starts. aws lambda-microvms create-microvm-image \ --name hello-microvm-test \ --code-artifact "uri=s3://my-bucket/artifact.zip" \ --base-image-arn arn:aws:lambda:us-east-1:aws:microvm-image:al2023-1 \ --build-role-arn arn:aws:iam::123456789:role/MicroVMBuildRole Image build took about 3 minutes. Once done: aw

Today I launched Beach Day API , a developer API for real-time beach, ocean, water quality, advisory, amenity, access, and condition data. The goal is simple: make it easier for developers to build apps and tools around beach conditions without having to manually gather data from scattered sources. Beach Day API currently supports beaches across the United States and Australia , and returns structured JSON that can be used in travel apps, weather apps, surf tools, tourism websites, hotel and resort platforms, map-based search experiences, local discovery apps, and coastal safety dashboards. What the API includes Beach Day API can provide data such as: Beach profiles GPS and location data Ocean and weather conditions Water quality grades Advisories and closures Amenities Access details Beach-specific safety and visitor information A proprietary Beach Day Score The Beach Day Score is designed to give developers a fast way to surface whether a beach looks like a good choice for visitors on a given day. Why I built it Most weather APIs are broad. They can tell you temperature, wind, rain, or general conditions, but they usually do not answer the real user question: “Is this a good beach day?” That question depends on more than weather. It can involve water quality, advisories, closures, ocean conditions, amenities, beach access, and the actual visitor experience. Beach Day API is built around that more specific use case. Example use cases Some things developers could build with it: A beach finder app A surf or coastal conditions app A hotel or resort beach conditions widget A local tourism guide A travel planning tool A map-based beach discovery experience A safety dashboard for advisories and closures A recommendation engine for nearby beaches Built for simple integration The API uses API-key authentication and returns clean JSON responses. I wanted it to be straightforward enough that a developer could start testing quickly and then build it into a real product withou

A growing trust crisis You throw a task at an AI, and seconds later it spits out a dozen-line SQL blob. You copy and paste it, and it runs. But do you really feel confident about it? That’s probably the daily reality for every data analyst and developer today. Today, AI can generate runnable SQL. The problem is that you don’t know whether you can trust it. Once queries involve deeply nested window functions and multi-level subqueries, they become difficult to review, debug, maintain, and port across databases. Research has shown that when LLMs lack sufficient schema context and domain knowledge, they can produce hallucinated outputs, such as incorrect table joins, flawed aggregation logic, or missing critical filters. According to dbt’s 2026 benchmark, even the most advanced LLMs achieved only 64.5% accuracy on Text2SQL tasks. In other words, one out of every three SQL queries generated by AI may contain errors. A 2025 Stack Overflow survey revealed a more unsettling fact: only 2.7% of professional developers place a high level of trust in AI tools. Meanwhile, 42% of submitted code is already AI-generated, and only 48% of is reviewed by humans. The latest GitHub data paints a worrying picture: AI is rapidly producing code, while humans are struggling to keep up with the pace of verification. Is this an AI problem, or a problem of how we approach it? If you think about it, the root of the problem may not lie in AI itself, but in the fact that we are asking AI to do things it is not well suited for. AI is well suited for brainstorming. It can help you clarify your thinking, break down complex requirements, and explore multiple possible solutions. However, when it comes to precise execution, the statistical nature of AI means it cannot guarantee a 100% correct result. Entrusting final code generation to a probabilistic model is, in itself, a methodological mismatch. The decline of Stack Overflow indirectly confirms this point. The number of new questions on this platfo

About a year ago, I turned my gaming PC into a local AI Lab. And yes, the most important word in that sentence is LOCAL . Let me tell you the story of how I sacrificed my gaming hours to build several tools, and now I'm going to tell you about this one that I use every single day. The Problem: Token bankruptcy Day to day, all of us developers who work with Artificial Intelligence share the same headache: tokens and rate limits . We're all victims of the high prices that come with constantly running inference with AI agents like Claude Code, Codex, or Gemini CLI (yeah, I love working from the terminal, I LOVE CLIs). While I was building AI systems (agent orchestration, LLM fine-tuning ), I was burning through way too many tokens. I tried tweaking the prompts and cleaning up the junk in my context, but the real devourer of my quota showed up when I had to learn a new tool. I was implementing solutions in QGIS (QGIS is a free, open-source Geographic Information System (GIS) software that allows users to create, edit, visualize, analyze, and publish geospatial data on maps) for a project and I didn't know the interface 100%. Like any dev facing something new, I leaned on AI agents: I'd take a screenshot, send it over, and ask for explanations. Here's an important fact that hurt my wallet: A screenshot on my MacBook (Full HD resolution of 1920x1080) burns about 258 tokens per tile on models like Claude. That adds up to roughly 1,548 tokens per image (sounds like a lot, and yeah my friend, it is way too much when we're talking about context). Now imagine sending dozens of these images a month trying to understand a complex interface as a 2x dev (99x, I'd say, in this new AI era). I was eating through my hourly Claude allowance just doing visual queries, leaving me with no quota left to generate the actual code I really needed for my development. The Epiphany (and the Hardware) One day, during a forced break thanks to a Claude rate limit , I looked over at my Gaming PC. I

You added a useLayoutEffect to measure a tooltip, shipped it, and the next time your Next.js (or Remix, or Gatsby) dev server rendered a page on the server, the console lit up: Warning: useLayoutEffect does nothing on the server, because its effect cannot be encoded into the server renderer's output format. This will lead to a mismatch between the initial, non-hydrated UI and the intended UI. To avoid this, useLayoutEffect should only be used in components that render exclusively on the client. The warning is correct, the suggested fix ("only use it on the client") is unhelpful, and the obvious workaround — just switch to useEffect — quietly reintroduces the visual bug you used useLayoutEffect to kill in the first place. useIsomorphicLayoutEffect is the small hook that resolves the standoff. This post explains exactly why the warning happens, why the two naive fixes are both wrong, and what the one-line hook actually does. Why useLayoutEffect Exists At All React gives you two effect hooks that look nearly identical: useEffect runs after the browser has painted. Its callback is queued and fires asynchronously once the frame is on screen. useLayoutEffect runs before the browser paints, synchronously, right after React has mutated the DOM but before the user sees anything. That timing difference is the whole point. If you need to read layout — getBoundingClientRect , scrollHeight , the measured width of a node — and then write a style based on it, you have to do it before paint. Otherwise the user sees one frame of the wrong layout, then a flicker as your useEffect corrects it. The canonical example is a tooltip that has to position itself relative to its own measured size: function Tooltip ({ targetRect , children }) { const ref = useRef < HTMLDivElement > ( null ); const [ pos , setPos ] = useState ({ top : 0 , left : 0 }); useLayoutEffect (() => { const { height , width } = ref . current ! . getBoundingClientRect (); // place the tooltip above the target, centered s

Why I Still Believe in Zero-Cost BFF Layers After 6 Months (And What Broke) Honestly, I didn't expect to be writing this article. Six months ago, I built capa-bff — a zero-cost BFF framework that won a hackathon gold medal — and I thought I had it all figured out. "This is perfect," I told myself. "Zero configuration, works with any Spring Boot app, solves all the frontend aggregation problems." Spoiler alert: It didn't. Don't get me wrong — it's still great for what it is. But here's the thing about building developer tools: the real world has a way of humbling you. Let me walk you through what I learned, what works, what doesn't, and who should actually use this thing. What Even Is a BFF Anyway? If you're new to the term, BFF stands for Backend For Frontend . It's that intermediate layer between your frontend clients (web, mobile, mini-programs) and your backend services. The idea is simple: instead of making the frontend stitch together data from multiple backend APIs, you have this middle layer that does it for you. ┌─────────────┐ ┌─────────────┐ ┌─────────────┐ │ Frontend │ -> │ BFF │ -> │ Backend │ │ (Web/Mobile)│ │ Aggregation │ │ Services │ └─────────────┘ └─────────────┘ └─────────────┘ The benefits are clear: Fewer network calls from the client Customized responses for each client type Better caching opportunities One place to handle auth/transformations But here's the catch most articles don't tell you: adding a BFF layer means another service to maintain , another deployment , another thing that can break . For small teams and startups, that cost can feel too high. That's exactly why I built capa-bff: I wanted a zero-cost BFF layer that you can just drop into your existing Spring Boot app. No new service, no extra deployment — just add the dependency and start aggregating APIs. How It Actually Works (Code Example) Let me show you the basics. With capa-bff, you define your aggregation in a simple annotation: @BffRoute ( path = "/user-dashboard" ) public

For years, scam prevention has leaned heavily on awareness. Be careful. Do not click suspicious links. Check the sender. Call the organisation directly. Do not trust urgent payment requests. Slow down before you act. These messages are useful, and they should not disappear. But awareness is no longer enough to describe what serious scam defence requires. The Scams Prevention Framework, or SPF, moves the conversation from “make users more careful” to “make the scam ecosystem harder to exploit.” That shift is important. Modern scams do not succeed only because a user failed to notice a warning sign. They succeed because scam operators move through gaps between messaging channels, platforms, brand impersonation, payment pressure, fake infrastructure, multilingual persuasion, reporting delays, and weak post-report disruption. Awareness helps at the point of decision. SPF requires capability across the whole chain. In my view, awareness alone covers about 28% of the real scam defence problem. The rest sits in evidence quality, intelligence sharing, infrastructure disruption, multilingual interpretation, safe financial harm context, recurrence monitoring, and operational response. That is why SPF should not be read as an education policy. It should be read as an operating model. The Awareness Ceiling Awareness is a front-line control, not a full defence system. It helps users recognise risk, but it cannot remove the fake page, connect related reports, preserve evidence, disrupt a fake app, identify a phone-linked abuse path, or monitor the next replacement domain. It also assumes the user has enough time, confidence, language support, and emotional distance to make a calm decision. Many scam situations are built specifically to remove those conditions. Scammers do not only trick uninformed people. They create urgency for informed people. They create authority for cautious people. They create routine-looking payment requests for busy people. They create private pressure fo

Claude Code is a genuinely powerful CLI coding agent. Its context window handling and multi-file reasoning set a high bar in 2026. But it comes with real constraints - it requires an Anthropic API key, charges per token, locks you into Claude models only, and its source code is closed. For developers running local-first workflows, working in air-gapped environments, or simply preferring auditable tooling, those limitations are dealbreakers. The good news: the open-source ecosystem has matured significantly. Nine production-ready alternatives now cover every major workflow pattern - from terminal-first pair programming to fully autonomous task execution. Why Open Source Matters for AI Coding Agents AI coding agents operate at a high level of system trust. They write files, run commands, and modify your repository. That makes transparency genuinely important - not just philosophically. Open-source licensing lets you read the code, audit its behavior, self-host without sending data to a third party, and customize it for your team's needs. Beyond trust, the practical advantages are real. Open-source agents are model-agnostic by design. They connect to whichever LLM you prefer - Claude, GPT, Gemini, DeepSeek, or a local model via Ollama - letting you optimize for cost and capability on a per-task basis rather than being locked to one pricing tier. OpenCode - The Closest Open-Source Drop-In for Claude Code OpenCode has emerged as the de facto open-source answer to Claude Code in 2026, crossing 161,000 GitHub stars under an MIT license. It connects to over 75 LLM providers via Models.dev - including local Ollama models - and lets you switch providers mid-session. Internally it uses a dual-agent architecture: a Plan agent handles task decomposition while a Build agent executes changes. LSP integration brings symbol resolution into the terminal. Multi-session support lets you run parallel agents on the same project simultaneously. OpenAI Codex CLI - Auditable and Sandbox-Fir

Every few months, I see another post declaring: "Prompt engineering is dead." Usually, the argument goes something like this: AI models are getting smarter. They understand natural language better. You no longer need carefully crafted prompts. On the surface, that sounds reasonable. But after building AI workflows and experimenting with modern frameworks, I think the opposite is happening. Prompt engineering isn't disappearing. It's evolving. And if you're building AI applications, not just chatting with AI, you'll probably rely on it more than ever. Prompt Engineering Was Never About Fancy Prompts One of the biggest misconceptions is that prompt engineering is about writing magical sentences that somehow unlock hidden AI capabilities. It isn't. Good prompt engineering is about giving an AI system exactly what it needs to complete a task reliably. Consider these two examples. Poor prompt: Write Python code. Better prompt: Write a Python FastAPI endpoint that accepts a CSV upload. Requirements: Use Python 3.12 Validate file type Handle exceptions Return JSON responses Include comments explaining each step The second prompt isn't "clever." It's simply clearer. And clarity scales. AI Models Are Better, But They Still Need Context Modern LLMs have become incredibly capable. They can: Generate code Explain algorithms Debug applications Write tests Refactor functions But they still don't know: Your architecture Your coding standards Your API contracts Your deployment strategy Your business requirements That information comes from you. And the way you provide it matters. Prompt engineering is fundamentally the practice of supplying useful context. Every AI Framework Depends on Good Prompts Take a look at the most popular AI frameworks. Whether you're using: LangChain LangGraph CrewAI LlamaIndex Every one of them eventually sends prompts to an LLM. Even sophisticated agent systems are built from sequences of prompts. Agents don't eliminate prompt engineering. They multiply

I had been running a small (3 people) software company for about 4 years. Since closing down, I recently hung out at a friend's company to see what they were working on (15 ppl). To preface: I'm a heavy user of Claude (rarely write code by hand), but what I'm seeing in person has been rather shocking to me, and I wanted to calibrate with others. In particular: - the code is not the source of truth anymore; it's ask claude to write, and ask claude to explain - LoC, abstractions, and all those "so

Grab's security team built Palana, a Kubernetes-native secure execution platform, to run autonomous AI agents safely. Unlike deterministic software, model-driven agents exhibit unpredictable tool-use, code-writing, and prompt injection risks. Palana contains these threats at the infrastructure level using isolated namespaces, out-of-process control planes, and proxy-mediated, Vault-backed secrets. By Patrick Farry

A new OpenAI research paper shows how AI agents are transforming work, enabling longer, more complex tasks and expanding productivity across roles.