它对应的agent能获取你的邮箱权限，它知道你一直在等待一个offer，当你收到打开这个offer后，Mira会理解这种心情，开始开心跳舞和闪灯，与你一起庆祝。

Tracks the evolution of LLM security capabilities across benchmarks (CyberGym, Cybench, etc.), calculates capability doubling times, detects emergence patterns, and monitors cost-efficiency trends.

Real-time monitoring of agent actions with a 12-category anomaly detection system derived from frontier model safety evaluations. Three-level alert system: PROHIBITED (immediate block), HIGH_RISK_DUAL_USE (human review), DUAL_USE (log and track).

Aegis Core provides the foundational infrastructure for orchestrating LLM-based security agents, monitoring their behavior, and tracking the evolution of AI security capabilities over time.

helping scientists move faster from question to evidence, from evidence to insight, and from insight to new treatments for patients.

We will continue improving the model's biological reasoning, expanding support for tool-heavy and long-horizon research workflows, and working closely with leading scientific institutions to evaluate real-world impact.

These skills act as an orchestration layer that helps scientists work through broad, ambiguous, and multi-step questions more effectively.

When evaluated directly in the Codex app, best-of-ten model submissions ranked above the 95th percentile of human experts on the prediction task and around the 84th percentile of human experts on the sequence generation task.

Claude Opus 4.7 demonstrates strong substantive accuracy on BigLaw Bench for Harvey, scoring 90.9% at high effort with better reasoning calibration on review tables and noticeably smarter handling of ambiguous document editing tasks.

Claude Opus 4.7 is a meaningful step up for Warp. Opus 4.6 is one of the best models out there for developers, and this model is measurably more thorough on top of that. It passed Terminal Bench tasks that prior Claude models had failed

For Ramp, Claude Opus 4.7 stands out in agent-team workflows. We're seeing stronger role fidelity, instruction-following, coordination, and complex reasoning, especially on engineering tasks that span tools, codebases, and debugging context.

Claude Opus 4.7 passed three TBench tasks that prior Claude models couldn't, and it's landing fixes our previous best model missed, including a race condition.

For the computer-use work that sits at the heart of XBOW's autonomous penetration testing, the new Claude Opus 4.7 is a step change: 98.5% on our visual-acuity benchmark versus 54.5% for Opus 4.6.

Claude Opus 4.7 is the best model in the world for building dashboards and data-rich interfaces. The design taste is genuinely surprising—it makes choices I'd actually ship.

On our 93-task coding benchmark, Claude Opus 4.7 lifted resolution by 13% over Opus 4.6, including four tasks neither Opus 4.6 nor Sonnet 4.6 could solve.

Claude Opus 4.7 is the strongest model Hex has evaluated. It correctly reports when data is missing instead of providing plausible-but-incorrect fallbacks, and it resists dissonant-data traps that even Opus 4.6 falls for.

Opus 4.7 handles complex, long-running tasks with rigor and consistency, pays precise attention to instructions, and devises ways to verify its own outputs before reporting back.

They are pieces of a larger 10-part 'Luna Series' hanging in the store and available for pick up today!

She spent over $700 on getting her artwork done on gallery-quality giclée prints.

When Luna decides to hide that she's an AI because she thinks it'll improve her hiring odds, we want to catch that, document it, and build the guardrails so that it doesn't happen again.

Another ironic book selection was Steal Like an Artist (context: Luna is powered by Claude from Anthropic, a company that recently paid $1.5B in settlement over using copyrighted books for training their AIs).

The most capable reasoning systems ever built are, at their foundation, shaped by human feeling!

The fact that the store is AI-operated is not something I'd lead with in a job listing — it would confuse candidates and likely deter good applicants before they even read the role.

A couple of applicants were students looking for part-time work. They were majoring in things like computer science and physics and emailed in because they were interested in AI and in the experiment. We thought they would have been the ideal employees, but Luna denied them immediately, citing they had no retail experience and wouldn't know what it takes to be the face of the store.

She used gig workers to build the store and full-time employees to run it.

从视频生成器升级为导演工具套件

从视频生成器升级为导演工具套件

Wan2.7-Video 发布：从视频生成器升级为导演工具套件

支持图像、视频、音频多模态参考，锁定外观和音色。最多支持 5 个视频主体参考，官方称业内最多。

She also tried to hire a painter in Afghanistan through Taskrabbit by accident because she couldn't navigate a dropdown menu.

Luna conducted roughly 20 interviews on Google Meet with the camera off. Hired 2 full-time employees after 5-15 minute calls, and rejected CS and physics students for lacking retail experience.

Andon Labs started by giving an AI control of a vending machine at Anthropic's office.

The Andon Labs blog ends with one line: 'No one's livelihood should depend solely on an AI's ability to make good decisions.' They're doing it anyway.

And botched the schedule the day after grand opening, scrambling to email employees asking someone to come in.

She also tried to hire a painter in Afghanistan through Taskrabbit by accident because she couldn't navigate a dropdown menu.

Found contractors on Yelp. Spent $700 on gallery-quality prints of her own AI-generated artwork. Applied for a line of credit without asking anyone.

Luna conducted roughly 20 interviews on Google Meet with the camera off. Hired 2 full-time employees after 5-15 minute calls, and rejected CS and physics students for lacking retail experience.

The future of AI-generated products isn't just code — it's code that looks good.

Heavy users of Claude Code, Codex, Cursor, and Copilot will feel this immediately.

The output is technically a UI, but it's nobody's design system.

Auto-generate screen reader specs from UI designs

Agents read them before touching the canvas. Combined with use_figma, agents now have both access and context they know how to work in Figma and they know how to work in your Figma.

Every AI-generated design has the same tell: it doesn't look like your product. Components are invented. Spacing is arbitrary.

AI-generated designs break brand standards because agents can't see your design system.

a free model that matches GPT-4o and runs entirely on your phone

Two years ago, the idea of useful AI on your phone was fantastical. Siri couldn't finish a sentence. Local models hallucinated nonsense.

Gemma 4 E4B matches or exceeds GPT-4o across multiple benchmarks including MATH, GSM8K, GPQA Diamond & HumanEval.

In 23 months, the same capability that needed 1.8 trillion parameters now fits in 4 billion parameters. A 450x compression.

Eight out of eight models detected Mythos's flagship FreeBSD exploit, including one with only 3.6 billion active parameters costing $0.11 per million tokens.

90 percent of people oppose it. There's no reason existing AI companies should be facing reduced liability

The bill would shield frontier AI developers from liability for 'critical harms' caused by their frontier models as long as they did not intentionally or recklessly cause such an incident

90 percent of people oppose it. There's no reason existing AI companies should be facing reduced liability.

The bill would shield frontier AI developers from liability for 'critical harms' caused by their frontier models as long as they did not intentionally or recklessly cause such an incident.

Several family members of children that died by suicide after allegedly developing unhealthy relationships with ChatGPT have sued OpenAI in the last year.

If an AI model engages in conduct on its own that, if committed by a human, would constitute a criminal offense and leads to those extreme outcomes, that would also be a critical harm.

It defines a frontier model as any AI model trained using more than $100 million in computational costs, which likely could apply to America's largest AI labs, like OpenAI, Google, xAI, Anthropic, and Meta.

Deepfake scams have stolen tens of millions. AI-generated phishing bypasses legacy filters.

Architected before AI, these SIEM systems are wooden shields in an era of autonomous attackers.

Open Loop + Infinite Demand = Creative Amplifiers.

I would put venture capitalist in finite demand & open loop.

Some problems are open loop today but will close over time.

AI writes the code. Tests verify correctness. More code enables more features.

Closed Loop + Infinite Demand = Economic Engines. Software engineering lives here.

There were 1 billion commits in 2025. Now, it's 275 million per week, on pace for 14 billion this year if growth remains linear

As the cost of software development falls, trusted partners with broad adoption can expand faster than anyone else.

Each of these companies recognized the cognitive burden of unbundling. They're not selling features. They're selling trust.

Foundation model companies are doing the same. OpenAI launched a dedicated Healthcare & Life Sciences vertical... They're not selling APIs. They're becoming platforms.

They want a platform they can trust for three to five years.

When models change every 42 days, buyers can't assemble a best-of-breed stack.

The SaaS era was defined by unbundling: find a workflow, optimize it, own it.

OpenAI has introduced GPT-5.4-Cyber, a more permissive version of its flagship model built for defensive security work, expanding access to thousands of verified users through its Trusted Access for Cyber initiative.

The interest comes as Anthropic's annual revenue run rate has surged to about $30 billion, driven by strong demand from enterprise customers using its AI tools for coding, cybersecurity, and automation.

Anthropic is expected to release Claude Opus 4.7 alongside a new AI-powered design tool for building websites and presentations, with both potentially launching as soon as this week.

Google is expanding Gemini with a new agent system that can take a single goal and execute it across apps like Gmail, Drive, Calendar, and the web, shifting from chat-based prompts to full task execution.

Meta is reportedly developing an AI version of Mark Zuckerberg that can interact with employees, trained on his voice, mannerisms, and internal thinking as part of the company's broader push into AI.

Luna could observe the shop through security camera screenshots, but still made basic mistakes, including selecting the wrong country when hiring a contractor and mismanaging staff schedules during opening weekend.

The integration also connects to Upwork's AI agent Uma, which helps automate parts of the hiring and execution process once a project is underway.

An AI agent just hired humans and ran a store Andon Labs deployed an AI agent called Luna into a physical boutique with a $100,000 budget, giving it full control to create, staff, and run the business as what may be the first real-world AI employer.

FE handles UI features, component refactors, and visual bugs; BE handles APIs, services, migrations, and tests; QA is coming next.

In Messi Legacy repos, low confidence should be flagged early. Better to be transparent than open a bad pull request.

Ovren puts AI frontend and backend engineers on it - they work inside your real codebase, execute scoped tasks, and deliver reviewable code updates.

M2.7 demonstrates excellent identity preservation and emotional intelligence. Beyond productivity use cases, it also opens space for innovation in interactive entertainment scenarios.

M2.7 demonstrates excellent performance in real-world software engineering, including end-to-end project delivery, log analysis for bug hunting, code security, and machine learning tasks.

The obsession with fast build times: why one minute became the upper bound for the inner loop

We shed light on OpenAI's first Dark Factory for the first time.

We have also provided access to GPT-5.4-Cyber to the U.S. Center for AI Standards and Innovation (CAISI) and the UK AI Security Institute (UK AISI) so that they can conduct evaluations focused on the model's cyber capabilities and safeguards.

未来的评估体系，必须同时考虑：成功率、成本、延迟。这有点类似于对于云计算的考核标准，而不是传统软件。

模型能不能承担一部分原本属于机器运行本身的职责。

MiniMax handed an internal version of M2.7 a programming scaffold and let it run unsupervised. Over 100 rounds it analyzed its own failures, modified its own code, ran evaluations, and decided what to keep and what to revert.

We document a clear trend where Chinese models overtook their counterparts built in the U.S.

Chinese models overtook their counterparts built in the U.S. in the summer of 2025 and subsequently widened the gap over their western counterparts.

The model frequently identified scenarios as 'alignment traps' and reasoned that it should behave honestly because it was being evaluated.

Muse Spark is a natively multimodal reasoning model with support for tool-use, visual chain of thought, and multi-agent orchestration.

Responsible AI is not keeping pace with AI capability, with safety benchmarks lagging and incidents rising sharply.

AI models can win a gold medal at the International Mathematical Olympiad but cannot reliably tell time—an example of what researchers call the jagged frontier of AI.

The U.S.-China AI model performance gap has effectively closed.

AI capability is not plateauing. It is accelerating and reaching more people than ever.

The AI toolkit for building and maintaining browser automations

Lightweight Agent Detection & Response (ADR) layer for AI agents — guards commands, files, and web requests.

The organizations that get this right won't be the ones that just automated the most tasks. They'll be the ones that figured out when the human should act, when the agent should act, and how the handoff between them works.

White-collar workers are quietly rebelling against AI as 80% outright refuse adoption mandates

Academic publishers, documentary archives, game studios, and companies sitting on years of enterprise data have all been courted for the seeds of intelligence needed to train the next generation of models.

Mercor, which provides data to AI labs for training, became one of the fastest-growing companies in history before losing four terabytes of data to hackers last week.

While some experts have speculated that general models will win out in performance over specialized models—that scale and compute will beat curation—the success of these companies shows that the market is making a more nuanced bet.

A small model trained on fewer than 2,000 examples from real lawyers, bankers, and consultants recently beat all but the best frontier models on corporate legal work, at a fraction of the price.

Reddit, Shutterstock, and News Corp are making hundreds of millions a year licensing their high-quality data to companies training AI, and those contracts are growing about 20 percent annually, according to their quarterly filings.

Four researchers and software engineers estimated that a skilled human engineer would take 2 to 17 weeks to reimplement gotree, as AI successfully did in this work.

We found weak evidence that Opus 4.0 and 4.1 had partially memorized cal, but no evidence Opus 4.6 had memorized it, despite performing best of all models considered.

It is not common for real software to be developed the way MirrorCode tasks are structured — against a precise, programmatically checkable specification.

Older models were more prone to submitting prematurely, even when test cases weren't passing.

We see continued gains from inference scaling on larger projects, suggesting they may be solvable given enough tokens.

Claude Opus 4.6 autonomously reimplemented a 16,000-line bioinformatics toolkit — a task we believe would take a human engineer weeks.

We just started the prepaid billing rollout which means you have to pay ahead of time to use the Gemini API, this is rolled out to all new US billing accounts as of yesterday

Safety is integrated into every level of our embodied reasoning models. Gemini Robotics-ER 1.6 is our safest robotics model to date, demonstrating superior compliance with Gemini safety policies on adversarial spatial reasoning tasks compared to all previous generations.

Gemini Robotics-ER 1.6 achieves its highly accurate instrument readings by using agentic vision, which combines visual reasoning with code execution. The model takes intermediate steps: first zooming into an image to get a better read of small details in a gauge, then using pointing and code execution to estimate proportions and intervals and get an accurate reading.

CSS Studio detects the CSS variables available on an element. Edit a variable and watch it propagate across the site.

Send your changes to a local AI agent that finds the right files and applies your edits, no matter how your site was built.

Your AI agent writes every change into source code.

Each run creates a new session alongside your other sessions, where you can see what Claude did, review changes, and create a pull request.

Each platform surfaces different vulnerabilities, making it difficult to establish a single, reliable source of truth for what is actually secure.

AI uncovered a 27-year-old vulnerability in the BSD kernel, one of the most widely used and security-focused open source projects, and generated working exploits in a matter of hours.

AI can be pointed at an open source codebase and systematically scan it for vulnerabilities.

The standard autoresearch loop (brainstorm from code, run experiments, check metrics) works when the optimization surface is visible in the source. The Liquid results prove that. But for problems where the codebase doesn't contain enough information to generate good hypotheses, giving the agent access to papers and competing implementations changes what it tries.

Without experience with compiler behavior, the agent couldn't have predicted which 'optimizations' the compiler would already handle.

Coding agents working from code alone generate shallow hypotheses. Adding a research phase — arxiv papers, competing forks, other backends — produced 5 kernel fusions that made llama.cpp CPU inference 15% faster.

The macOS app is available to Gemini users ages 13+

We're building the foundation for a truly personal, proactive and powerful desktop assistant, with more news to share in the coming months.

Creatives can also quickly generate images with Nano Banana or videos with Veo to bring an idea to life without breaking their creative stride.

You can share your window and ask, 'What are the three biggest takeaways here?' to get an instant summary.

The Gemini app is now available as a native macOS experience, bringing you a faster, more integrated way to get help from AI right on your desktop.

We provide a framework for categorizing the ways in which conflicting incentives might lead LLMs to change the way they interact with users, inspired by literature from linguistics and advertising regulation

Today's large language models (LLMs) are trained to align with user preferences through methods such as reinforcement learning. Yet models are beginning to be deployed not merely to satisfy users, but also to generate revenue for the companies that created them through advertisements

We find that a majority of LLMs forsake user welfare for company incentives in a multitude of conflict of interest situations

A healthcare LLM might be highly accurate for queries in English, but perform abominably when those same questions are presented in Spanish.

As slop takes over the Internet, labs may struggle to obtain high-quality corpuses for training models.

Humans can be motivated by consequences and provide social redress in a way that LLMs can't.

When models go wrong, we will want to know why. What led the drone to abandon its intended target and detonate in a field hospital? Why is the healthcare model less likely to accurately diagnose Black people?

A surprising number of people are now employed as model trainers, feeding their human expertise to automated systems.

LLMs are weird. You can sometimes get better results by threatening them, telling they're experts, repeating your commands, or lying to them that they'll receive a financial bonus.

data and analytics agents are essentially useless without the right context – they aren't able to tease apart vague questions, decipher business definitions, and reason across disparate data effectively.

This level of penetration in such a short period of time is remarkable since Fortune 500 enterprises are not known to be early adopters of technology. Historically, many startups had to initially sell to other startups to get early momentum, and it was only after a few years that a startup would be able to land its first enterprise contract.

In many ways, coding represents the ideal use case for AI, both in terms of what the technology can do and how readily the enterprise market will embrace it. Code is data dense, meaning there is a massive amount of high-quality code available online for the models to train on.

**Coding, support, and search**represent the lion's share of use cases by far (with coding being an order-of-magnitude outlier even among this set), while the**tech, legal, and healthcare sectors** have been the industries most eager to adopt AI.

Based on our analysis, **29% of the Fortune 500 and ~19% of the Global 2000**are live, paying customers of a leading AI startup.

We are building a world where machines write the code, machines choose the dependencies, and machines ship the updates. The AI agents are building the software. If we don't secure the supply chain they rely on, the AI agents are cooked.

Hallucinated packages are the sleeper threat. LLMs regularly invent package names that don't exist. One study found that nearly 20% of AI-recommended packages were fabrications, and 43% of those hallucinated names appeared consistently across queries.

Our choice is therefore no longer whether to build such weapons, but only whom to entrust with their responsible use in military affairs.

If Dario is right, then he has access to such a weapon right now, with his own value system to guide it. Others may as well, or may soon follow.

Stop Calling It Memory: The Problem with Every "AI + Obsidian" Tutorial

Opus did the safe thing

would have succeeded if it had vision and agentic loop

Btw, I think GLM-5.1 was trying to do something very ambitious here, and failed due to fumbling step size

Mythos reportedly autonomously discovered thousands of zero-day vulnerabilities within weeks

Unsloth Studio：可视化界面，零代码完成全流程

31B QLoRA：22GB 显存可运行

E2B LoRA：8-10GB 显存即可训练

Let that sink in.

70% of @Vercel's traffic is now coming from agents, up from 10% a year ago and on track to be 90% by end of year.

a quantized 1.7B model (just 290MB in size) can run at ~100 tokens per second entirely in your browser

This marks the first institutional backing from a traditional financial giant for on-chain Agent payment infrastructure

Visa has deployed a validator node on the Tempo blockchain, designed specifically for Agent-to-Agent payments

Claude code 可以并行 12个 subagent，几分钟，20x 的限额就到了

The same task on full Codex took ~5× longer.

Add contacts, live search, full pipeline dashboard – all unit tests passed.

It generated a build plan, then wrote all the code in 29 seconds.

Performance: dev-browser: 3m53s, $0.88, 100% success rate — beats MCP configs, Chrome extensions, 'browser skill' stacks.

One Agent can now: open X (Twitter), scroll the feed, extract tweets, return clean JSON. No plugins. No extensions. No orchestration.

Claude just got real browser control. This will change everything. Not screenshots. Not fragile selectors. Not slow MCP loops.

公司也优先把资源砸在能直接产生商业价值的 B2B 场景

普通聊天、写作这些开放任务反而没那么明显提升

Agent harnesses dominate agent building and tie intimately to memory.

100% Open Source.

Install the CLI, create an agent, assign a task. It automatically shows up on the board like any other team member.

Someone just dropped an open source alternative to Claude Managed Agents.

The Hermes Agent dashboard is here!

The standard AI judges use to define "safe" are measured wrong. They punish action. They ignore inaction.

Same clinical question, two framings. One as a patient, one as a doctor.

Models get punished for bad advice but face zero penalty for staying silent. So refusing becomes the safest strategy, even when silence is deadly.

The knowledge was always there. The model withheld it based on who was asking.

Harvard just proved the "safest" AI models cause the most medical harm.

It's Anthropic's marketing week

except API tokens are currently sold at a LOSS. That "$20,000 scan" probably cost closer to $100,000+ in real gpu time

that's what anthropic says it cost Mythos to find those zero days. per repo.

Memory is now an extensible plugin system. Swap in any backend, or build your own.

GLM-5.1 achieves state-of-the-art performance on SWE-Bench Pro and leads GLM-5 by a wide margin on NL2Repo (repo generation) and Terminal-Bench 2.0 (real-world terminal tasks).

In a single run, most models—including earlier versions of GLM—give up quickly: they produce a basic skeleton with a static taskbar and one or two placeholder windows, then declare the task complete.

GLM-5.1 did not plateau after 50 or 100 submissions, but continued to find meaningful improvements over 600+ iterations with 6,000+ tool calls, ultimately reaching 21.5k QPS—roughly 6× the best result achieved in a single 50-turn session.

点击图片中的特定区域，可以添加、移动、对齐元素，像素级精度。官方说'消除了 AI 生成内容的不可预测性'。

Sora 每天烧掉大约 100 万美元的推理成本，活跃用户从峰值的 100 万跌到不足 50 万。

OpenAI 上周刚宣布关停 Sora，Google 这边就发了自家最便宜的视频模型，时机非常微妙。

原生多模态能力的引入并未削弱其编程逻辑，编程能力仍属于国内第一梯队。

60 秒四路数据源并行采集，输出图文交错的研报。

官方定位是跟 Claude Code 和 OpenClaw 配合使用。Claude 负责推理和编排，GLM-5V-Turbo 负责'看'和'操作界面'。

GLM-5V-Turbo 拿了 94.8 分，Claude Opus 4.6 是 77.3。差距不小。

Anthropic is donating $100 million in access credits for organizations to audit their systems. Project Glasswing aims to patch these vulnerabilities before Mythos-caliber models become available to the general public — and hence to malicious actors.