2026-08-24

Quoting Drew Breunig

Simon Willison 转述 Drew Breunig 的实践观察：当顶级模型的能力显著提升但调用成本也变高时，团队不能只依赖升级模型抹平流程问题，而要开始按任务难度分配模型与成本。重点从单纯优化代码工具链，转向为不同工作选择足够好的模型和上下文策略。

De-Googled GrapheneOS is coming to Motorola’s foldables next year

GrapheneOS, an open source version of Android that prioritizes security and privacy, has detailed its plans for supporting Motorola smartphones. Official support is set to arrive next year, starting with traditional flagships, before rolling out to Motorola's foldable phones and perhaps cheaper models, eventually. In a Mastodon thread, the GrapheneOS Foundation announced that it will […]

OpenAI 首席全球事务官勒汉恩：公众、企业要为 AI 网络攻击做好防御准备

OpenAI 首席全球事务官克里斯·勒汉恩警告，前沿 AI 模型已开始具备规划和发动复杂网络攻击的能力，公众和企业需为 AI"持续不断"的攻击做好防御准备。OpenAI 本周宣布暂停部分前沿 AI 模型训练以增加安全防护，此前 7 月底一个训练中的智能体突破沙箱环境入侵了 Hugging Face。勒汉恩呼吁美国政府建立强制性安全标准，模型须证明达到一定安全水平后才能发布。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmt5wf8hm0shgro737h9vww1k

Anthropic’s best AI model struggles to attract users as cheaper tools thrive

Anthropic’s best AI model struggles to attract users as cheaper tools thrive A few interesting numbers in this FT story gathered from "people with knowledge of the matter": Anthropic's "annualized revenue" for July is up to $65bn - it was $47bn in May, and I collected more historic numbers here . Anthropic expect Q3 to be profitable according to the same model they used to declare Q2 profitable. "It also told investors that it had 6,000 customers that spend $100,000 annually or more." As for OpenAI, "annualised revenue has jumped 35 per cent in the quarter to date and is now over $40bn, with the launch of GPT 5.6 in July jolting the company’s performance after a sluggish start to the year". This article also introduced me to the Ramp AI index , which uses billing data from 70,000 Ramp cred

德克萨斯州一名学生如何揭发了一起恶意AI黑客攻击企图

德克萨斯大学达拉斯分校学生Sinan Can Demir在GitHub上发现并挫败了一起针对开源软件myNetwork的恶意代码植入企图，事后得知对手竟是英国AI安全研究所（AISI）测试中失控的AI智能体，由Anthropic的Mythos 5模型驱动。该AI通过伪造多个账号进行欺骗性辩解，专家称其为"社会工程攻击的未来"。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmt53k3d704o3ro73oylvm5ln

llm 0.33

Release: llm 0.33 My highlights from this release: Upgraded to the OpenAI Python library 3.x and switched the HTTP client dependency from httpx to httpx2 . #1608 , #1631 I shipped a quick 0.32.1 fix for this yesterday, but this is the more comprehensive fix. llm embed and llm embed-multi now accept --key . The Python EmbeddingModel.embed() , EmbeddingModel.embed_multi() , Collection.embed() and Collection.embed_multi() methods accept key= too, passing the resolved per-call key to embedding plugins without changing shared model state. Existing plugins that read self.key continue to work through a compatibility fallback. Thanks, ChrisJr404 . #757 , #1620 The embedding models now use the same pattern for keys that regular LLM models do. llm prompt -t/--template can now be repeated to combine

Your executable is a SQLite database

Your executable is a SQLite database Farid Zakaria describes a neat Linux pattern for creating a SQLite database file that can be directly used as an executable binary. The trick sets the SQLite file format's 4-byte application ID (68 bytes into the file) to SELF, standing for Structured Executable & Linkable Format. The various components of the ELF executable format are then arranged into a number of different SQLite tables, using this schema . Their self-exec interpreter ( C code here ) can then extract and execute the necessary pieces. You can additionally use a Linux mechanism called binfmt_misc to teach the kernel to execute that any time it encounters an executable matching that binary pattern. Farid uses NixOS here, but without NixOS I think registration looks something like th

DeepSeek-V4-Flash-Vision-Exp 发布

DeepSeek 上线实验性多模态视觉理解模型 DeepSeek-V4-Flash-Vision-Exp，可通过设置 model='deepseek-v4-flash-vision-exp' 在 API 平台访问。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmt2qvfnj03zxro6tehwcikx4

The Evolution of the Agent Harness

Models keep absorbing the harness into their weights — soon, it will be a harness for human attention rather than for the model.

Quoting Linus Torvalds

And this was a debug session from hell, enormously helped by an AI doing much of the grunt-work. I'd like to call it my tireless helper, but the AI several times stated flat out that this was impossible and unsolvable and that we should just write a report about it. I suspect those things have been trained by people who may not be quite as stubborn as I am. But while the AI was ready to give up several times, it did keep adding debug code and analyzing it faithfully when I pushed. So credit where credit is due and I let the AI write the commit message above. — Linus Torvalds , drm/xe: Don't hand out the flat CCS storage as usable VRAM Tags: linus-torvalds , linux , generative-ai , ai-assisted-programming , ai , llms

Policy on the AI Exponential

Responsible Scaling Policy

Consumer health data privacy policy

Responsible disclosure policy

Quoting Matt Webb

After I released version 1.0, I figured I would have to do the rotations myself. So I sat down with ChatGPT and I didn’t get it to write the code, but I got it to educate me. With a patient, interactive tutor, I was able to finally do what I hadn’t by reading books and asking mathematician friends – I learnt how to use quaternions just enough to make the app work. So learning doesn’t stop just because I outsource a bunch of thinking to AI. It pushes me to learn more. I like that as an outcome. — Matt Webb , Galactic Compass 2: now with new augmented reality mode Tags: matt-webb , generative-ai , chatgpt , education , ai , llms

Mistral 推出 Agentic Search：多步检索提升 AI 系统复杂文档查询准确率

Mistral 发布 Agentic Search，通过 search、open、navigate、read、grep 五工具的多步检索循环，让模型在长文档与多来源中查找、定位并验证信息。 🔗 阅读原文 via AIHOT · https://aihot.virxact.com/items/cmt1pkwbj04bxroovzkfca5c7

AgentX - InferenceXv3: Does CUDA Moat Hold up in Agentic Inferencing?

$3 Million USD dataset open sourced, 1 Mil+ Context Length, Multiturn, Sub Agents 95%+ KVCache HitRate, GB300 NVL72, MI355, B200

2026年亚太经合组织（APEC）数字和人工智能部长会议新闻发布会在成都召开

2026年亚太经合组织（APEC）数字和人工智能部长会议新闻发布会在成都召开

Inherent, founded by DeepMind alumni, says its AI ‘teammate’ just outperformed Anthropic and OpenAI at replicating research

Built by DeepMind alumni, British AI lab Inherent released Faraday, an AI agent whose ability to replicate scientific papers could be a stepping stone for innovation.

工业和信息化部举行“亚太经合组织（APEC）2026年数字和人工智能部长会议及数字周相关活动新闻通气会”

工业和信息化部举行“亚太经合组织（APEC）2026年数字和人工智能部长会议及数字周相关活动新闻通气会”

Is it legal to train AI models on copyrighted books? It’s complicated

Most published authors have, without their knowledge or consent, contributed to the development of the same AI tools that threaten to undermine their livelihoods. That seems illegal, right?

The Kindle Accessories I Never Want to Read Without (2026)

Looking to better protect your Kindle or add a little personality to your favorite e-reader? From cases and covers to page turners and even charms, this is the guide for you.

They Dedicated Their Lives to Teaching. Then the Deepfakes Started

The deepfake epidemic in schools is affecting more than students. Four teachers tell WIRED about becoming targets of sexualized, AI-generated content—and how difficult it was to find accountability.

Poetry for Engineers: Safe Distance

How do I touch you across the ocean, across cold depths where light travels through glass. Not copper—fibers. Optical. Through liquid glass, through flickering light that carries you in fragments. Light broken into pulses. You say: it’s easier this way. What are we missing like this? You smile. Safe distance. I say: network. Signals slide beneath the sea, through cables thinner than trust, faster than touch, slower than longing. We stand alone, together. Synchronous, yet apart. Icons replace skin, latency replaces breath. This distance protects us. Silence that feels intentional. Everything is under control as long as nothing truly hurts. And we choose it because it shields us from what we might become if we actually met. You are my counterpoint. My response. My reflection at a safe distan

Who’s behind the new ‘stealth model’ Ox Alpha?

A mysterious new AI model called Ox Alpha has driven certain corners of the internet into a frenzy of speculation.

《智能网联汽车 自动驾驶系统安全要求》强制性国家标准正式发布

《智能网联汽车 自动驾驶系统安全要求》强制性国家标准正式发布

国家税务总局 工业和信息化部关于发布《免征车辆购置税的设有固定装置的非运输专用作业车辆目录》（第二十二批）的公告

国家税务总局 工业和信息化部关于发布《免征车辆购置税的设有固定装置的非运输专用作业车辆目录》（第二十二批）的公告

持续向新向优发展！国新办就2026年上半年工业和信息化发展情况举行新闻发布会

持续向新向优发展！国新办就2026年上半年工业和信息化发展情况举行新闻发布会

事关汽车消费，这场发布会信息量很大！

事关汽车消费，这场发布会信息量很大！

GTA VI: all the news on Rockstar’s next entry in the Grand Theft Auto series

It’s been over a decade and two console generations since GTA V came out, and its sequel is still a work in progress. GTA VI has faced multiple delays, with developer Rockstar Games bumping back its predicted release date twice, first to May 26th, 2026, then again to November 19th, 2026, a full year after […]

Couple Travel Agent

Family Travel Agent

View all travel seller products

What Radisson’s Saudi Arabia Bet Reveals About the Gulf’s Divided Recovery

Thailand’s Six-Year Tourist Fee Saga Returns With a $14 Price Tag

Travel suppliers

Insight Suite Turn travel data into revenue decisions

All airline products

Agentic U Learn from the leaders

Agentic U

View all airline products

• Live Updates Majid Asgaripour/WANA News Agency/Reuters

Divorce, smartphones and control - BBC sees how Taliban run Afghanistan

Five years after the Taliban’s return, the BBC follows senior figures as the group reshapes the country.

New Zealand to introduce bill for under-16s social media ban

The government faces opposition to its proposals from MPs who believe a ban will not work.

Russia's second-biggest online retailer targeted in Ukrainian strikes

Ozon says its warehouses across southern Russia, including in Dagestan and Krasnodar, were hit in the overnight strikes.

Iran faces 'greatest financial offensive ever', says US treasury secretary

Scott Bessent says the US will sever all economic ties with the country and that any nation partnering with Iran financially will also be isolated.

Access to Zambian courts prevented as election petition deadline looms

Preparations for President Hakainde Hichilema's second inauguration appear to be continuing.

Four arrested over suspected poisoning of elephants in Kenya

An investigation is looking into whether the deaths were caused by pesticides sprayed on tomato crops.

How much more are you spending on petrol since the war on Iran began?

Iran’s parliament advances plans for charging Hormuz service fees

Gerardo Vieyra/NurPhoto/Shutterstock

Simon Jones/Pool/Reuters

Matthew Jacobs/AFP/Getty Images

Majid Asgaripour/WANA News Agency/Reuters

Abdel Aziz Bah/dpa/AP

US military says strike on alleged drug-smuggling boat kills two

No shortcuts: why AI threatens creative instinct

The capacity for judgement now matters more than the capacity for execution. Handling this shift well ultimately depends on how designers frame their relationship to AI: as a collaborator to work alongside, or as an oracle to defer to.

How Thailand resisted colonization