标签：#ai

This weekend, AgentTrust ID went live in production. As of today, all five SDKs are published: pip install agenttrustid npm install @agenttrustid/sdkgo get github.com/agenttrustid/sdk/go cargo add agenttrustid # Maven / Gradle # id.agenttrust:agenttrustid:0.3.0 The SDKs are open source under Apache 2.0 at github.com/agenttrustid/sdk . The hosted platform is running at app.agenttrust.id in a controlled beta. Why I built this AI agents broke the assumptions that machine-to-machine security was built on. An API key answers one question: who is calling. It asks it once, at the door. An agent decides its next action at runtime, from context nobody wrote by hand. The same agent that summarized a document a second ago might now try to email it, delete it, or chain a task to another agent. A credential that only proves identity has no opinion about any of that. Agents need a decision at the action boundary : should this specific action happen, right now, on whose behalf . Answered at runtime, every time, with an audit trail and a kill switch. What's running Everything below is live in production today, not a roadmap: Per-action authorization. Every consequential action passes a pre-flight check. The Guardian pipeline routes each action by risk: deterministic rule checks for the common path, a policy engine for mutations, and AI-backed review for destructive operations. Fail-closed where it counts. Opaque, instantly revocable tokens. Credentials are at_ references with no standing authority of their own . The server decides on every use, so revocation is one call, effective immediately. Scoped delegation. When one agent hands work to another, the grant narrows instead of copying : subset scopes, independent TTLs, independently revocable, bounded chain depth. Read-only sessions with time-boxed elevation. Sessions start safe and rise only on approval, for a bounded window, then revert on their own. One model across surfaces. MCP tools, agent-to-agent calls, and direct API inte

https://www.youtube.com/watch?v=VD9zEKQEJxo 这视频深入拆解了人工智能强化学习之父、图灵奖得主理查德·萨顿（Richard Sutton） 在2026年5月共同发表的一篇仅有7页、零算法、零跑分的哲学立场论文。这篇论文提出了 “行动认知 AI”（Enactive Artificial Intelligence，简称 Enactive AI）的概念，并在科技界和资本圈引发了巨大震动（甚至让红杉、英伟达、谷歌联合下注了11亿美元成立新公司）。 视频从 核心概念、哲学脉络、理论内在矛盾、认知科学质疑 以及 产业界的三路对赌 五个维度，极其详细地复盘了视频的所有核心内容： 一、 什么是“行动认知 AI”（Enactive AI）？ 视频强调，全网很多地方都把 Enactive （行动认知/生成认知）和 Generative （生成式 AI，如 GPT、Sora）混淆了，但两者的底层逻辑恰恰相反 [ 00:50 ]： 生成式 AI（Generative AI）： 核心是 续写和预测 。通过已有画面或文本，被动地去预测下一帧、下一个词长什么样 [ 01:07 ]。 行动认知 AI（Enactive AI）： 核心是 在互动中现生成认知 。认知不是大脑被动接收信号并建立静态世界模型，而是“你动了手，世界才向你显现” [ 01:47 ]。 > 举例： 人去拿杯子，不是眼睛先拍下一张静态照片让大脑去死算距离、角度 [ 01:53 ]，而是手往前探的过程中，随着角度、光影的实时动态变化，杯子的形状和可抓取性才在动作里一点点“长出来” [ 01:59 ]。 感知和行动硬死在一起，无法拆分。 这套理论源自认知科学中的 自创生（Autopoiesis）与自主性（Autonomy） [ 02:21 ]。它认为智能体应该像生物一样自我维持、组织，由内在生存需求去塑造感知，而不是一个干等着外部指令输入输出的机器 [ 02:24 ]。 二、 萨顿为什么要发这篇哲学论文？ 萨顿并不是一时性起，这是他为了对抗当前“大模型路线”打出的最后一张哲学底牌： 2019年《苦涩的教训》： 主张人类手写规则干不过堆算力、让机器自己学的通用方法 [ 02:47 ]。 2024年《大世界假设》： 真实世界远比静态内部模型复杂，智能体必须在运行中实时学习 [ 02:59 ]。 2025年《经验时代》： 人类数据是有限的，AI 必须靠自己生成自己的经验长大的 [ 03:12 ]。 2025年9月： 直指整个 AI 行业走错路，大模型堆数据去超智是死路一条 [ 03:19 ]。 这篇论文补上了最后一把火： 之前的论证全是算力、数据和复杂度的“机械账” [ 03:25 ]。而这一次，他第一次把强化学习（RL） 和 认知科学（行动认知）接在了一起，从本体论层面证明： 大模型路走不通，认识世界这件事本身，就只能通过行动和互动的经验来发生 [ 03:39 ]。 为此，2026年初论文共作者创办了 Ineffable Intelligence 公司，号称要造出完全不需要人类数据、靠自己学习的 AI，直接拿到了红杉、英伟达、谷歌 11 亿美元的巨额融资（估值 51 亿美元） [ 03:55 ]。 三、 论文隐藏的两大致命致命逻辑“回旋镖” 视频话锋一转，指出萨顿借来的这套哲学地基里，埋着两根砸中他自己的“大柱子”： 柱子 1：砸中了萨顿的“奖励假设”（自相矛盾） [ 04:35 ] 强化学习的号称教条： 奖励假设（Reward Hypothesis），即所有目标、意图都可以写成“最大化外部给定的标量分数” [ 04:53 ]。David Silver 甚至喊出“奖励就够了” [ 05:13 ]。 行动认知哲学的教条： 自主性（Autonomy），即什么是好坏、成败，标准必须从智能体随时会散架的“物理组织和生存危机”中自发长出来，不能由外部权威操控 [ 05:27 ]。 裂缝： 标准强化学习的奖励函数（Reward Function）是人类设计者用代码硬塞进去的（他律） [ 05:55 ]；而生物判断好坏是为了顶住熵增、维持结构不崩（自主） [ 06:11 ]。论文里作者自己也承认：强化学习的评估标准依然由外部奖励定义 [ 06:38 ]。 内驱动机能救场吗？ 比如好奇心驱动或求知驱动。视频认为不能，因为诸如“优化预测误差”的总结优化目标，依然是人类在架构层死死规定好的，根本不是智能体出于生存忧关的自发需求。没有真正的生命威胁，就没真正的意义生成 [ 07:12 ]。 柱子 2：砸中了萨顿自己的《苦涩的教训》 [ 07:49 ] 萨顿当年痛骂：研究者总忍不住把人类以为的思考结构（比如语法树、手工特征检测器）硬塞进 AI 架构里，这长期必被碾压 [ 08:13

Here's where you can watch Christopher Nolan's 'The Odyssey' in jaw-dropping IMAX 70mm.

Today I’m talking with Mustafa Suleyman, the CEO of Microsoft AI. And I’m actually going to keep today’s intro short — I’m working from my wife’s family farm this week, as you’ll see in the video, but also this is a real burner of an episode. We covered everything from Mustafa’s approach to training new […]

Summer Game Fest may be over, but the flood of gaming news isn't. Nintendo just announced that it'll be holding its next Direct showcase event on June 9th at 10AM ET. The stream, which you can find here, will be followed by a Treehouse Live event, and it sounds like it'll be pretty lengthy. "The […]

Two years ago, when Microsoft first revealed that it was bringing four Xbox-exclusive games to the PS5 and Nintendo Switch, it made the announcement far more complicated than necessary. That's not likely to improve anytime soon. In fact, things now seem more confusing than ever as the company tries to appease both fans and the […]

OpenAI to recast hit chatbot as a route to higher-margin products before a potential IPO.

I've said one and mean another, and I've used one when I needed another. Please bear with me as I note the similarities and differences between scroll-driven animations, scroll-triggered animations, container query scroll states, and view transitions for my future self. Scroll-Driven, Scroll-Triggered, Scroll States, and View Transitions originally handwritten and published with love on CSS-Tricks . You should really get the newsletter as well.

Tech Companies Regret Firing Engineers for AI: The Quiet Rehiring Nobody's Talking About [2026] Klarna's CEO Sebastian Siemiatkowski stood on stage in 2024 and bragged that AI had replaced 700 customer service employees. The stock market loved it. LinkedIn influencers celebrated. And then, quietly, in 2025, Klarna started hiring humans again. That single reversal tells you everything about why tech companies regret firing engineers for AI. I've watched this pattern unfold across the industry, and a viral YouTube video by Pooja Dutt documenting these failures is now pulling over 10,000 views per day. The audience isn't just curious. They're vindicated. The tech industry laid off over 260,000 workers in 2023 alone, according to Layoffs.fyi , with many companies explicitly citing AI automation as justification. Now, in 2026, the bills are coming due. The companies that swung hardest at the "AI replaces engineers" thesis are the ones scrambling hardest to undo the damage. Why Did Companies Fire Engineers for AI in the First Place? The logic seemed airtight. AI can generate code faster than humans. AI can handle customer queries at scale. AI doesn't need benefits, PTO, or performance reviews. Executives saw a clean line from "AI generates output" to "we need fewer people," and they drew it with a Sharpie. I've been in enough executive planning meetings to know exactly how this plays out. Someone demos an AI tool that produces a working prototype in 20 minutes. The room gets excited. The CFO asks how many engineers they can cut. Nobody asks the harder question: what happens when that prototype needs to survive contact with production? The answer is that it breaks. Badly. Klarna is the poster child, but they're far from alone. Apple has spent two full years struggling with AI-driven improvements to Siri, despite being one of the most well-resourced engineering organizations on the planet. Even with virtually unlimited budget and talent, replacing deep engineering expertise

When we build something complex—whether it’s a skyscraper, a gourmet meal, or a piece of software—we usually start with a plan. In software development, however, it’s easy to skip that step. We often jump straight into implementation, focusing on how to write the code instead of the intent behind it. Over time, this leads to rework, confusion, and systems that don't quite match our original goals. Spec-Driven Development (SDD) is an approach that shifts the focus back to the plan. Instead of starting with code, you start with a Specification : a clear, structured description of what the software should do. You then use an AI coding agent as a high-speed collaborator to help turn that specification into working code. 🔍 What is a “Spec”? A Specification (or “Spec”) is a written contract between your intention and the final product. It isn't a 50-page manual; it's a living document that defines: What the system should do. How it should behave in different scenarios. Which constraints and rules it must follow. From Prompts to Specifications There is a massive difference between a vague prompt and a structured spec. Loose prompts often lead to inconsistent results and "hallucinations," whereas clear specifications give the AI a much better target to hit. Bad Prompt: > “Build me a login system.” Good Spec: A good spec provides the clarity an AI (or a human) needs to succeed. You don’t need a 10-page document to benefit from specs; you need clarity, not length. 🛠️ Example Spec: Login Endpoint Overview Allow users to log in using email and password. Endpoint POST /api/login Request { "email" : "user@example.com" , "password" : "string" } Behavior Success: If email and password are correct → return a token and user info. Invalid Credentials: If credentials don't match → return INVALID_CREDENTIALS . Invalid Input: If fields are empty or the email format is wrong → return INVALID_INPUT . Rules Passwords must be stored hashed (e.g., bcrypt). Token expires in 24 hours. Security:

"# The AI Cost Crisis: How Startups Can Survive the Tokenpocalypse\n\n## Introduction\n\nThe artificial intelligence boom has brought unprecedented innovation, but it has also ushered in a era of spiraling costs. Training state-of-the-art models now requires millions of dollars in compute resources, while simultaneously, the cryptocurrency token market shows signs of a potential collapse—a \"Tokenpocalypse.\" For AI startups, this dual crisis presents an existential threat: how to sustain innovation when both traditional funding avenues and speculative token economies are under pressure? This post explores practical strategies for AI startups to navigate this landscape, focusing on cost optimization, alternative funding, and strategic pivots that can turn crisis into opportunity.\n\n## Understanding the Cost Explosion\n\n### The Compute Crunch\n\nModern AI models, particularly large language models (LLMs) and multimodal systems, demand vast computational resources. Training a single cutting-edge model can consume exaflops of processing power, translating to cloud bills that easily exceed $10 million for a single training run. For startups without deep-pocketed backers, these costs are prohibitive.\n\n### The Token Market Volatility\n\nParallel to the AI boom, the cryptocurrency space experienced explosive growth through token launches—initial coin offerings (ICOs), decentralized finance (DeFi) tokens, and utility tokens for AI-driven projects. However, regulatory crackdowns, market saturation, and declining investor sentiment have led to a sharp downturn. Many tokens have lost significant value, and launching new tokens has become increasingly difficult, removing a once-viable funding path for AI startups.\n\n## Strategies for Survival\n\n### 1. Embrace Model Efficiency\n\nInstead of chasing ever-larger models, startups can focus on efficiency techniques that deliver comparable performance at a fraction of the cost:\n\n- Model Distillation : Train smaller \"student\

Google says Gemma 4 12B is "designed to bring agentic, multimodal intelligence directly to your laptop", further noting that the new model can be combined with Google AI Edge to "build and experiment locally, on everyday machines". This integration allows for a wide range of capabilities, from autonomous data processing to generating visual insights and even building webpages or executing tools. By Sergio De Simone

InfoQ celebrates its 20th anniversary. To mark the occasion, we have published a walk-through of the trends InfoQ called early, where they sit on the adoption curve today, and how that curve may evolve over the next decade. By InfoQ

Machine learning has its limits—how is it being used?

Birgitta Böckeler, Distinguished Engineer at Thoughtworks, returns to discuss the rapid evolution of AI in software delivery. She touches on the evolution from vibe coding, the changing tools landscape and the more autonomous agents that, besides higher velocity, introduce higher risk. By Birgitta Böckeler

In this article, the author examines how AI is transforming phishing from a manual, targeted activity into an automated and scalable attack model. The article breaks down each stage of the phishing lifecycle, showing how AI improves reconnaissance, profiling, content generation, delivery, and interaction, while outlining layered defenses that combine controls, processes, and user awareness. By Marco Rizzi

Apple's biggest event of the year is nearly here. The company's Worldwide Developers Conference will spotlight updates to iOS, macOS, and all of Apple's other operating systems, and this year's event could also include a major overhaul for Siri. Here's how you can watch along live. When WWDC will happen and where you can watch […]

A compliance team at a mid-size bank in Abu Dhabi processes new customer applications every week....

While vibecoding, you sometimes need some background music. But music can also be a massive...

I ran a 4-bit medical-triage model on a laptop GPU and on a CPU. For one patient, the GPU said urgent and the CPU said emergency. Same model file, same prompt, same input. Here's the mechanism and why "validated on hardware X" doesn't mean what you'd hope. I've been building Aegis-MD , a local-first emergency-department triage console. You hand it a structured clinical picture: chief complaint, vitals, age, pain score, a few risk modifiers, and it returns an urgency category on the Australasian Triage Scale (ATS 1–5), where ATS-1 means resuscitate now and ATS-5 means this can wait two hours . The whole thing runs on-device: a quantized MedGemma 4B served through Ollama, a small RAG layer over open guidelines, and a deterministic rule-based floor underneath the model. I never set out to write about floating-point arithmetic. But while running my evaluation set across two machines, I hit a result that stopped me, and the explanation turned out to be more interesting and more current than the textbook answer most people reach for. The setup, and why a 4-bit model Two things about Aegis-MD's design matter for this story. First, it's local by design. Triage data is about as sensitive as data gets, so nothing leaves the machine. The trade-off is that I'm running a small, heavily quantized model: MedGemma 1.5 4B at Q4_K_XL , about 3.4 GB rather than a frontier API. Four-bit weights are the price of running offline on consumer hardware. Second, I tested on two configurations on purpose. The intended deployment is local GPU inference (an RTX 5070 Ti Mobile, 12 GB). But the public demo runs CPU-only on Cloud Run, because GPU instances need a paid quota I don't have. So I ran the same evaluation against both: the GPU build and the CPU build, same model, same code, same prompts. The eval is 17 hand-written cases spanning all five ATS levels, cardiac arrest down to a medical-certificate request. (Seventeen is a smoke test, not a validation; I won't quote a percentage off a sampl