标签：#AI

Focus: Unit Tests ONLY — no integration, no E2E Stacks: Node.js (NestJS/Express) · React.js · Python · Angular · Laravel Goal: AI generates unit tests consistently, deterministically, without hallucination IDE: Cursor (Primary) + Claude (Secondary) Part 1 — Best Single Library Per Stack (Final Decision) Do not mix libraries. Pick one per stack, configure it fully, never deviate. | Stack | Library | Why This One | |---|---|---| | Node.js / NestJS / Express | Jest | Native DI mocking, @nestjs/testing built around it, widest ecosystem | | React.js | Vitest + @testing-library/react | Native Vite/ESM support, Jest-compatible API, 3–10x faster | | Python | pytest | De facto standard, fixture system eliminates boilerplate, best plugin ecosystem | | Angular | Jest (replace Karma) | Karma is deprecated in Angular 17+; Jest is the official migration target | | Laravel | Pest | Modern syntax, built on PHPUnit, higher signal-to-noise ratio | Rule: If someone suggests a second library for the same stack, reject it. One library per stack, configured once, followed always. Part 2 — IDE: Cursor (Only Choice for This Goal) Why Cursor and Not VS Code / WebStorm | Capability | Cursor | VS Code + Copilot | WebStorm | |---|---|---|---| | Project-level AI rules | ✅ .cursor/rules/ | ❌ | ❌ | | Codebase-aware context | ✅ @codebase | Partial | Partial | | Run terminal + read output | ✅ Composer | ❌ | ❌ | | Multi-file generation | ✅ Agent mode | Limited | ❌ | | Custom instructions per filetype | ✅ | ❌ | ❌ | | MCP server integration | ✅ | ❌ | ❌ | Cursor's .cursor/rules/ system is the only IDE-native mechanism that injects persistent, project-scoped instructions into every AI interaction — this is what prevents hallucination at the source. Cursor Setup for This Project project-root/ ├── .cursor/ │ └── rules/ │ ├── unit-test-global.mdc ← applies to all files │ ├── unit-test-nestjs.mdc ← applies to *.service.ts, *.guard.ts │ ├── unit-test-react.mdc ← applies to *.tsx, *.component.tsx │ ├── unit-t

Your retriever returned the right documents. The similarity scores look fine. The answer is still wrong. If you've shipped RAG, you've seen this — and it's the failure that survives every retrieval upgrade. What everyone tries Reranker. Higher top-k. Hybrid search. A better embedding model. All of these chase the same goal: documents more similar to the query. They help when the right document wasn't being retrieved. They do nothing when the right document was retrieved and the answer is still wrong. Why it doesn't work Similarity answers "is this chunk about the same topic?" It does not answer "does this chunk contain the facts needed to support the answer?" Those come apart constantly. A chunk can be highly similar — same vocabulary, same subject — and contain nothing that actually grounds the answer. Hand the model a pile of on-topic text and it will produce a fluent, plausible, even cited-looking answer. The grounding is cosmetic: the text was nearby, not load-bearing. High similarity with a wrong answer isn't a contradiction. You asked retrieval to find related text. It did. Nobody asked whether the text was enough. The one shift Stop treating retrieval output as evidence. Treat it as candidate material that has to pass an explicit evidence check before it can support an answer. Put a step between retrieval and generation: does the retrieved set actually contain the facts this answer requires? If not, abstain. When the documents don't contain the facts, the system should return nothing rather than a confident guess. Relevant context in, only sufficient evidence allowed through. That's the line between a RAG demo and a RAG system you can trust in production. I write about the three boundaries where production RAG dies — query, evidence, output — from the angle of shipping under security and model constraints. Read the full version on my blog , where this connects to the practical RAG Failure Diagnosis Kit for teams debugging production RAG.

A few months ago, I was overwhelmed by everything happening in AI. Every week there was a new coding assistant, a new workflow, or someone claiming they built an app in just a few hours. It felt like if you weren't keeping up, you'd be left behind. I tried almost everything. Cursor. ChatGPT. Claude Code. Lovable. At first, I kept switching between tools, hoping one of them would magically make me a better developer. It didn't. The biggest lesson I learned wasn't about choosing the best AI tool. It was learning how to work with AI. These days, I don't start by asking AI to write code. I start by explaining the problem. I describe the feature, the business requirements, the edge cases, and what I want the final result to look like. Sometimes I ask ChatGPT to help me plan the implementation first. Once everything is clear, I pass that plan to an agentic coding assistant and start building. That one change made a huge difference. I spend less time writing boilerplate and more time thinking about architecture, user experience, and solving the actual problem. AI still gets things wrong, so I review everything before it goes into production. But instead of writing every single line myself, I'm guiding the process. Looking back, the first few months were the hardest. Now it just feels normal. The tools will keep changing, but I think the real skill is learning how to communicate with AI and use it as part of your development process. That's something worth investing in.

How to Access 50+ Chinese AI Models Through One API The Chinese AI ecosystem exploded in 2025-2026. DeepSeek dropped training costs by an order of magnitude. Qwen 3 ships 19 variants from 0.6B to 235B parameters. GLM-5 competes head-to-head with GPT-5 at 3% of the price. There's Kylin, Yi-Lightning, Hunyuan-T1, MiniMax-M1, Step-2-16K, and 40+ more models from a dozen labs. The models are incredible. The fragmentation is not. Every lab has its own API. Different auth headers. Different response formats. Different streaming protocols. Different error codes. If you wanted to try 5 models from 5 Chinese labs last year, you'd need 5 SDKs and 5 billing dashboards. Nobody has time for that. This is exactly the problem AIWave was built to solve. One API Key. 50+ Models. Zero Code Changes. AIWave is a unified API gateway that aggregates 50+ Chinese AI models behind a single endpoint. It speaks the OpenAI API format, which means every existing tool, SDK, and codebase in your stack works without modification. Here's what that looks like in practice: from openai import OpenAI # Point to AIWave instead of OpenAI client = OpenAI ( base_url = " https://api.aiwave.live/v1 " , api_key = " sk-your-aiwave-key " ) # Use DeepSeek V4 Pro response = client . chat . completions . create ( model = " deepseek-v4-pro " , messages = [{ " role " : " user " , " content " : " Explain MoE architecture " }] ) # Switch to GLM-5 — change one string response = client . chat . completions . create ( model = " glm-5 " , messages = [{ " role " : " user " , " content " : " Explain MoE architecture " }] ) # Try Qwen 3 235B — same thing response = client . chat . completions . create ( model = " qwen3-235b " , messages = [{ " role " : " user " , " content " : " Explain MoE architecture " }] ) That's it. Whatever you're already using — the OpenAI Python SDK, LangChain, LlamaIndex, Vercel AI SDK, a custom fetch wrapper — continues to work. You change the base URL and the model name, and suddenly you have acce

OpenAI’s Bonnie Xu discusses Kepler, an internal AI data analyst agent built to query 600+ petabytes of data. She explains how they overcome context window limits using MCP, automated code crawling, and RAG. Xu also shares how their team leverages scoped semantic memory for self-learning and utilizes AST-based LLM grading to build a robust, regression-free evaluation pipeline. By Bonnie Xu

CircleCI has launched Chunk Sidecars, a new capability designed to bring CI-style validation directly into an AI coding agent's inner development loop By Craig Risi

When Colin Farrell was doing press for the first season of the detective series Sugar, he had to be very careful with how he spoke. Sugar is a story about a quirky private detective, but it's also secretly a work of science fiction, something that doesn't become clear until halfway through the season. "I knew […]

On June 9th, Anthropic released its Fable generative AI model. Three days later, the US government classified it as a dangerous munition, and used its export-control authority to prohibit any foreign nationals from accessing it. Unable to differentiate between Americans and foreigners, the company shut off access for everyone. The government’s actions won’t help . The problem isn’t any one particular model; it’s the general trend of increasing AI capabilities. And any real solution requires the sort of collective action that just isn’t possible right now...

The crypto platform claims you can “pay anyone to do anything,” from quitting a job on camera to getting a memecoin-themed tattoo. But it mostly seems like people trying to scam each other.

LLM costs scale linearly with usage. A system processing 10,000 requests a day at $0.01 per request costs $100 daily — $365 a year. At enterprise scale, that's over $10,000. Cost optimization isn't about cutting corners. It's about spending tokens where they matter. Every token you waste is a token you could have spent on a better answer. Token budgeting The simplest way to control costs is to set limits. Per session, per task, or per day. Strategy 1: Per-Session Budgets Per-session budgets are straightforward: class SessionBudget : def __init__ ( self , budget_tokens : int = 10000 ): self . budget = budget_tokens self . used = 0 def allocate ( self , tokens : int ) -> bool : if self . used + tokens <= self . budget : self . used += tokens return True return False def remaining ( self ) -> int : return self . budget - self . used Strategy 2: Per-Task Budgets Per-task budgets are more useful. Different tasks need different amounts of context: task_budgets : classify : max_tokens : 100 model : qwen2.5-1.5b summarize : max_tokens : 500 model : qwen2.5-7b code_review : max_tokens : 2000 model : qwen2.5-coder-7b reason : max_tokens : 4000 model : qwen2.5-32b Strategy 3: Adaptive Budgets Adaptive budgets adjust based on what actually happens. If classification tasks consistently use 80 tokens, stop allocating 100: class AdaptiveBudget : def __init__ ( self ): self . task_history = {} def allocate ( self , task_type : str ) -> int : if task_type in self . task_history : return int ( self . task_history [ task_type ] * 1.5 ) return 1000 def record ( self , task_type : str , tokens_used : int ): if task_type not in self . task_history : self . task_history [ task_type ] = tokens_used else : self . task_history [ task_type ] = ( 0.9 * self . task_history [ task_type ] + 0.1 * tokens_used ) The exponential moving average (0.9 weight) means recent usage matters more than history. Adjust the weight based on how volatile your workloads are. API vs local inference Local inference

LLMs are unpredictable. They hallucinate, leak data, generate harmful content, or refuse legitimate requests. Guardrails constrain model behavior without sacrificing capability. The key is knowing which guardrails matter and which are just noise. Guardrails aren't about controlling the model. They're about controlling the risk. Input validation The most important guardrail. Bad input gets bad output, and bad input can also prompt-inject your system. Strategy 1: Prompt Sanitization Sanitize dangerous patterns early: import re class PromptSanitizer : def __init__ ( self ): self . dangerous_patterns = [ r " ignore\s+previous\s+instructions " , r " system\s+prompt " , r " you\s+are\s+now\s+free " , r " break\s+out\s+of " , ] def sanitize ( self , prompt : str ) -> str : for pattern in self . dangerous_patterns : prompt = re . sub ( pattern , " [REDACTED] " , prompt , flags = re . IGNORECASE ) return prompt This isn't bulletproof. Adversarial inputs are creative. But it catches the obvious ones, and the obvious ones are the most common. Strategy 2: Input Length Limits Length limits prevent token waste and timeouts: class InputValidator : def __init__ ( self , max_length : int = 10000 ): self . max_length = max_length def validate ( self , prompt : str ) -> tuple [ bool , str ]: if len ( prompt ) > self . max_length : return False , f " Input too long: { len ( prompt ) } > { self . max_length } " return True , " OK " Strategy 3: Content Filtering Content filtering blocks policy violations. The patterns here depend on your domain: class ContentFilter : def __init__ ( self ): self . blocked_topics = [ " violence " , " hate speech " , " self-harm " , " sexual content " , " illegal activities " , ] def filter ( self , prompt : str ) -> tuple [ bool , str ]: prompt_lower = prompt . lower () for topic in self . blocked_topics : if topic in prompt_lower : return False , f " Blocked: { topic } " return True , " OK " Simple string matching is fast but imprecise. For production, us

Running a 70B parameter model to summarize a 200-word email is wasteful. Running a 3B model to review production code is reckless. Most systems live somewhere in between — and that's where model routing comes in. It matches task complexity to model capability. The tradeoffs are real, but the savings are too. The routing problem People usually start with one model and stick with it. That works until you notice the cost, or the latency, or both. The alternative is building a router — something that decides which model handles which request. Four strategies work in practice: Capability-based — route by what the model can do Cost-aware — route by what you're willing to spend Latency-aware — route by how fast you need it Hybrid — combine them Each optimizes something different. Picking one is usually a decision about what hurts most. Capability-based routing The simplest approach. Classify the task, send it to the model that handles it. Task Model size Examples Classification, tagging 1-3B Qwen2.5-1.5B, Gemma-2-2B Summarization, extraction 3-7B Qwen2.5-7B, Llama-3.1-8B Code generation 7-14B Qwen2.5-Coder-7B, DeepSeek-Coder-V2 Complex reasoning 14-32B Qwen2.5-32B, Llama-3.1-70B Creative writing, analysis 32B+ Qwen2.5-72B, Claude, GPT-4 If the task doesn't need the bigger model, don't use it. A 1.5B model handles sentiment classification fine. It just won't write a coherent essay. Implementation is straightforward: ROUTING_RULES = { " classify " : { " model " : " qwen2.5-1.5b " , " max_tokens " : 100 }, " summarize " : { " model " : " qwen2.5-7b " , " max_tokens " : 500 }, " code_review " : { " model " : " qwen2.5-coder-7b " , " max_tokens " : 2000 }, " reason " : { " model " : " qwen2.5-32b " , " max_tokens " : 4000 }, " creative " : { " model " : " claude-sonnet-4 " , " max_tokens " : 8000 }, } def route_request ( task_type : str ) -> dict : return ROUTING_RULES . get ( task_type , ROUTING_RULES [ " reason " ]) The catch is classification itself. If you get the task type

Single-model systems are simple. Multi-model systems are powerful. The challenge isn't choosing models — it's designing the architecture that orchestrates them. A multi-model system isn't about having more models. It's about having the right model for the right task at the right time. Architecture patterns Five patterns cover most use cases: Pattern Complexity When to use Tradeoff Single Model Lowest Prototyping, simple tasks Limited capability Sequential Low Multi-step workflows Higher latency Parallel Medium Independent tasks Higher cost Hierarchical High Complex reasoning Complex orchestration Ensemble Highest Critical decisions Highest cost Pick the simplest one that works. Complexity is real, and it compounds. Sequential architecture Process tasks through a chain of models, each specializing in a step. Pattern 1: Pipeline Pipeline pattern — each model's output feeds the next: class ModelPipeline : def __init__ ( self ): self . models = [ { " model " : " qwen2.5-1.5b " , " task " : " classify " }, { " model " : " qwen2.5-7b " , " task " : " extract " }, { " model " : " qwen2.5-32b " , " task " : " reason " }, ] def process ( self , input : str ) -> str : current = input for model_config in self . models : current = self . call_model ( model_config [ " model " ], self . create_prompt ( model_config [ " task " ], current ) ) return current Latency adds up. Three models in sequence means three times the latency. Only use this when each step actually needs a different model. Pattern 2: Router Router pattern — classify the task, route to the specialist: class ModelRouter : def __init__ ( self ): self . classifier = " qwen2.5-1.5b " self . specialists = { " code " : " qwen2.5-coder-7b " , " math " : " qwen2.5-32b " , " creative " : " claude-sonnet-4 " , " general " : " qwen2.5-7b " , } def route ( self , prompt : str ) -> str : task_type = self . classify ( prompt ) model = self . specialists . get ( task_type , self . specialists [ " general " ]) return self . call_m

Hello Dev Community! 👋 It is officially Day 45 of my non-stop run toward full-stack engineering! Yesterday, I learned how to serve static HTML pages using Express routing. Today, I took a major step toward building premium, clean, and minimalist UI/UX styles by installing and mastering Tailwind CSS via the Tailwind CLI ! Instead of writing massive, chaotic external CSS stylesheet files, today I shifted to the industry-standard utility-first workflow to speed up my design iterations. 🧠 Key Learnings From Day 45 (Tailwind CSS Architecture) Tailwind doesn't give you pre-built components; it gives you atomic utility classes that let you build completely custom, high-end layouts directly inside your HTML structure. Here is the engineering breakdown: 1. Tailwind CLI Installation & Initialization I learned how to integrate Tailwind from scratch using npm packages instead of lazy CDN links. Installed the tailwindcss compiler core ( npm install -D tailwindcss ). Initialized the configuration hub using npx tailwindcss init . 2. Crafting the Content Map inside tailwind.config.js Understood how Tailwind optimizes final file weights using tree-shaking. I configured the structural template paths so the compiler knows exactly which files to scan for dynamic styling classes: javascript /** @type {import('tailwindcss').Config} */ module.exports = { content: ["./public/**/*.{html,js}"], // Scanning static folders cleanly theme: { extend: {}, }, plugins: [], }

"Email verification" sounds like one thing, but it's really a stack of checks of increasing depth and cost. Knowing what each layer actually proves helps you pick the right level instead of overpaying for verification you don't need. Layer 1: syntax The cheapest check: does the string look like a valid email address? A pragmatic regex catches obvious garbage ( asdf , a@@b , trailing spaces). It's instant and free, but weak on its own: nobody@asdf.asdf passes syntax and can't receive a single message. Layer 2: domain and MX records Next, does the domain actually accept mail? Every domain that receives email publishes MX (mail exchanger) records in DNS pointing to its mail servers. A quick DNS lookup tells you whether any exist. No MX (and no fallback A record) means the domain can't receive mail, so the address is undeliverable no matter how it's spelled. This single step removes a large class of fakes and dead domains. Layer 3: SMTP mailbox check The deepest level connects to the domain's mail server and begins the motions of sending a message to ask whether that specific mailbox exists, without actually delivering anything. It's the only layer that can hint a particular inbox is real, but it comes with real caveats: It's slow (a live connection per address). Many servers are "accept-all" and say yes to everything, so the answer is often meaningless. Lots of providers block or throttle these probes, and outbound port 25 is blocked on most modern hosting, so it's frequently unavailable anyway. SMTP checks matter most for cleaning old, cold lists, and far less for stopping junk at signup. The heuristics layer Alongside those, useful verification adds signal that has nothing to do with deliverability per se: Disposable detection: is it a throwaway provider? Role detection: is it info@ or admin@ rather than a person? Typo suggestions: "did you mean gmail.com?" for gmial.com . A deliverability score: one 0–100 number that rolls it all up so you can just threshold on it.

PSYONIC's prosthetic touch data is now training ABB robots. Gatik signed the first Fortune 50 commercial autonomous freight contract with PepsiCo. Burro drove Physical AI onto the construction site. Experts set $20k as the humanoid price target. And someone just called Edge AI the Windows of robotics. This week, Physical AI crossed three invisible lines at once. A company that makes prosthetic hands figured out that the touch data from amputees is exactly what industrial robots need to learn how to grip. A Fortune 50 company signed not a pilot but a commercial contract for autonomous freight. A 44-horsepower robot drove off the warehouse floor and onto the construction site. And two separate conversations about software and pricing suggest that the next wave of robotics adoption will be driven by access, not capability. Here is what happened, and why it matters beyond the headlines. Value Description Fortune 50 PepsiCo becomes first to sign a commercial contract for autonomous freight with Gatik $20k Target price point for humanoid robots, Robotics Summit consensus: achievable by 2028–2030 1M hours Burro's field experience backing the Grande 44 autonomous outdoor platform 100+ Pressure sensors per fingertip in PSYONIC's Ability Hand, now training ABB GoFa A Prosthetic Hand Is Now Teaching an Industrial Robot How to Grip The standard approach to teaching a robot how to handle objects has been simulation, teleoperation, or labor-intensive physical demonstrations. PSYONIC and ABB just introduced a different source of data : the hands of people who have already learned to feel again. PSYONIC's Ability Hand is a prosthetic with more than 100 pressure sensors per fingertip . The company has been collecting kinesthetic data from users with upper-limb amputations. That data, which captures how a human hand adjusts grip pressure, contact area, and force across thousands of everyday tasks, is now being fed as training data into ABB GoFa robot arm models. The implication is no

How I Cut My Multimodal AI Costs by 97% — A Freelancer's Guide Last month I almost killed a side gig because of a single line item on an invoice. A client wanted me to build a document-processing tool that could read scanned PDFs, pull text out of photos, and answer questions about charts. Easy enough — except I'd quoted the job assuming I'd use GPT-4o for the vision work. When I actually ran the numbers, I realized the API bill would eat my entire margin. I'd be working for free. Maybe worse. So I did what every freelancer does when the big-name vendor gets too expensive: I went hunting. And I landed on Global API, which routes to a bunch of multimodal models I've honestly never heard clients talk about. After a few weeks of testing, I figured out which ones are worth my billable hours and which ones aren't. This is everything I learned, plus the exact code I'm shipping to clients. Why Multimodal Even Matters for Solo Devs Two years ago, "multimodal" was a buzzword you'd hear at conferences. In 2026 it's table stakes. I've personally used vision models to: OCR receipts for an expense-tracking app (boring but pays the rent) Convert screenshots of legacy code into editable source for a Y2K-era company migration Read bar charts from PDF reports for a finance client who hates spreadsheets Analyze medical imaging samples for a startup MVP (this one was scary) Every one of those jobs started as a quick conversation with a prospect and turned into real invoices because I could say yes. The bottleneck was never capability — it was always cost. When GPT-4o charges north of $10/M output tokens, a single 2,000-token response on a tricky chart costs me about two cents. Multiply by 10,000 images per month and you've got a $200 API line item before you've paid yourself. That's a problem when the whole job is worth $400. So I tested every multimodal model I could find on Global API. Here's the lineup I ended up evaluating. The Contenders Nine models, three providers, one freelanc

Azure Functions shipped a serverless agents runtime in public preview at Build 2026. Agents are defined in .agent.md markdown files with YAML triggers, MCP server access, 1,400+ connectors, and sandboxed execution. The Functions team confirmed to InfoQ that the runtime adds no cold start overhead and no billing premium beyond standard Flex Consumption. By Steef-Jan Wiggers

In a new Windows Developer Blog post titled "Windows platform security for AI agents", Microsoft positions Windows as the trustworthy operating system for autonomous agents and introduces the Microsoft Execution Containers (MXC) SDK as the core of that strategy. The post argues that containment, identity and manageability must be built into the operating system. By Matt Saunders

There's a commercial logic that cuts against the idea that ASML would risk its export license to arm a Chinese customer.