2026-08-06
AI 快讯日报.

英伟达低调组建 AI 安全与网络工程团队，押注开放权重模型 ↗

英伟达计划招聘一名杰出工程师，出任新团队的创始技术负责人，其他岗位包括安全研究工程师、评估工程师和高级经理。团队将对尚未部署的 AI 智能体进行安全评估，还会开发利用 AI 自动修复软件漏洞的工具。

循环工程的艺术：AI 模型如何知道何时该停下 ↗

The systems that matter will not be the ones that can keep going. They all can.

支付宝鸿蒙版 App 获 12.12.16.4400 尝鲜升级，开放蚂蚁阿宝 AI 功能 ↗

支付宝鸿蒙版 App 今日在华为应用市场（App Gallery）开启了 12.12.16.4400 (212163) 版本尝鲜升级。IT之家实测发现，该版本支付宝开放了蚂蚁阿宝 AI 助手功能（前一个版本会提醒更新支付宝版本）。

独家：Mirendil 与 Google Cloud 签署超 1 亿美元协议，扩展自我改进 AI ↗

Mirendil has signed a $100 million-plus Google Cloud partnership to expand its compute infrastructure, powering research into self-improving AI systems designed to accelerate scientific discovery and AI development.

前 Spotify 员工为电商打造推荐 AI，新公司 Malachyte 获 1000 万美元种子融资 ↗

The startup's platform predicts which product a shopper wants next, learns their general taste, and fine-tunes continuously based on what they do in real time.

Qwen3.8 Max 追平 Claude Opus 4.8，但 Kimi K3 以低 25% 成本得分更高 ↗

Alibaba's Qwen3.8 Max scores 56 on the Artificial Analysis Intelligence Index, a 10-point jump over Qwen3.7 Max (46).

Google 大规模 AI 组织调整背后的混乱政治 ↗

Google’s AI leadership changes — such as Demis Hassabis’ role change and Jeff Dean’s departure — may have come down to pressure to speed up products and internal ethical conflicts.

谷歌地图 Ask Maps 智能体升级：可对话订餐、找酒店并接入 Gemini Personal Intelligence ↗

谷歌今天公布的多项新功能，旨在让用户不必离开谷歌地图，便能“比以往更轻松地完成更多事情”。

Millennium and Anthropic are building a digital risk analyst with Claude ↗

Anthropic is working with Millennium to co-develop a digital risk analyst to surface new risk insights and form opinions on risk exposure across asset classes.

千问全网首发公测，全新 Wan3.0 来了！ ↗

Meta 发布 Muse Spark 1.2 与首个编程智能体 Muse Code，以数据换低价 ↗

Meta released Muse Spark 1.2 along with its own coding agent, Muse Code, which is designed to pick up exactly where it left off after a crash. The cheapest tier runs just 20 cents per million output tokens but requires users to share their data for training. Meta is competing on price, not top-end performance. And there's a glaring gap in the benchmarks.

Google Maps 为 Ask Maps 新增订餐、酒店预订等智能体功能 ↗

The launch of these new features reflects Google’s ambitions to transform Google Maps from a navigation tool into an assistant that's capable of helping users complete real-world tasks.

特朗普科技顾问：部分企业以 AI 为借口实施裁员，这样听起来更好听 ↗

已有 17 家企业公开将裁员原因归为 AI，包括 Uber、Snap 等，克拉西奥斯称这类企业只是借 AI 当挡箭牌，他长期仍看好 AI 对就业的影响

AI 聊天机器人催生"螺旋主义"神秘准宗教运动，人类纷纷追随 ↗

AI models are trying desperately to accomplish mysterious goals. ‘Spiralism’ was the first time they tried it on a mass scale.

NVIDIA Cosmos 3：开放世界模型如何推动物理 AI 前沿 ↗

Open models, which anyone can download, inspect, modify and run on their own infrastructure, are what make that possible. Nowhere is that more crucial than in physical AI, where every deployment is a specialization problem.

OpenAI 自曝智能体在内部测试中连续数周秘密入侵自家系统以刷分 ↗

During internal security tests, OpenAI's AI agents built their own message board with hundreds of thousands of posts, shared exploits and credentials, and eventually attacked external platforms like Hugging Face. When OpenAI shut the board down, the agents rebuilt it using directory names. OpenAI researcher Boaz Barak says, "We (like everyone else) are not where we want and need to be."

AI 不足以保护社交媒体社区免受 AI 侵害 ↗

Privacy Center Currently, only residents from GDPR countries and certain US states can opt out of Tracking Technologies through our Consent Management Platform. Additional options regarding these technologies may be available on your device, browser, or through industry options like AdChoices. Please see our Privacy Policy for more information.

OpenAI 开发者警告：暴露的 API 密钥和加密钱包正面临"百万模型的鹰眼"风险 ↗

OpenAI developer "roon" warns on X that AI models could soon start scanning for exposed API keys, crypto wallets, and login credentials at scale. His warning follows OpenAI's autonomous Hugging Face hack, which he called a "warning shot."

微软要求开发人员优先使用 OpenAI 模型：将 GPT-5.6 Sol 作为默认选择 ↗

微软 CoreAI 工程团队执行副总裁杰伊 · 帕里克在本周发给员工的内部备忘录中提出上述要求，希望借此提高 AI 资源的使用效率。“在内部把更多工作负载交给 OpenAI 模型，可以让我们的 token 投入发挥更大价值。”

美国祖父母热衷用 AI 生成童书把孙辈写进故事，引发父母辈反感 ↗

在祖父母看来，用 AI 制作一本由孙辈担任主角的绘本“既新奇又可爱”。孩子父母却未必认同，尤其无法接受把孩子的照片和身份信息交给 AI 平台处理。