安全新基准：前沿模型已能自主开发浏览器漏洞
原文摘要：Researchers at Carnegie Mellon University built a new 评测 that measures how far AI agents can go when exploiting real vulnerabilities in Google's V8 engine. Mythos l 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
AI 每日快讯
AI 每日快讯
AI 产品、模型、开源工具和官方动态的时间流。保留历史记录，按分类、日期和标签继续筛选。
919历史快讯
60开源工具
8当前结果
05 月 16 日
2026-05-16 快讯
YouTube 向创作者开放深度伪造人脸检测
原文摘要：YouTube is opening its Likeness Detection tool to all creators 18 and older. The system spots AI-generated face fakes in other users' videos and lets creators file removal 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
AI 视频生成评测：画面更强但世界推理仍弱
原文摘要：A new 评测 called WorldReasonBench tests video generators not on image quality, but on physical and logical plausibility. ByteDance's Seedance 2.0 leads the field ahe 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
OpenAI 收购语音克隆创业公司，语音能力继续加码
原文摘要：OpenAI has acquired Weights.gg, a small startup that let users create and share AI voice clones of celebrities like Taylor Swift and Donald Trump. The team of around six n 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
百个 AI Agent 协作写码：OpenClaw 展示研发自动化成本
原文摘要：A three-person team led by Peter Steinberger keeps about 100 Codex instances running for the open-source project OpenClaw, driving OpenAI API spend to $1.3 million a month 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
The Decoder：AI made a tiny slice of Silicon Valley filthy rich and left the rest wondering why they both…
原文摘要：According to Menlo Ventures partner Deedy Das, about 10,000 people in Silicon Valley have amassed fortunes of over $20 million thanks to the AI boom at Anthropic, OpenAI, 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
The Decoder：Google says GEO and AEO are a myth and traditional SEO is all you need for AI search
原文摘要：Google says the SEO industry's favorite new buzzwords, "generative engine optimization" and "answer engine optimization," are just regular SEO by another name. In new docu 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。
The Decoder：Researchers train AI model that hits near-full performance with just 12.5 percent of its exp…
原文摘要：Researchers at the Allen Institute for AI and UC Berkeley have built EMO, a mixture-of-experts model whose experts specialize in content domains instead of word types. Tha 来源：The Decoder。建议继续查看原文，重点核对它影响的工具入口、成本、风险和真实使用场景。