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。

3231历史快讯
177开源工具
27当前结果
08 月 25 日 2026-08-25 快讯
MarkTechPost 官方资讯

MarkTechPost:Liquid AI Open-Sources Pipette: A Reproducible Benchmarking Suite That Measures On-Device Mo…

原文摘要:Model cards report quality under server-class, full-precision conditions. Those numbers rarely predict how the same model behaves on a phone. This week, Liquid AI released Pipette. 来源:MarkTechPost。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。

HarnessRouter 社区版发布:一个 API 统一接入多个 AI 编码 Agent

一句话结论:HarnessRouter 社区版以 Apache-2.0 协议开源,提供统一接口来运行 Codex、Claude Code、Hermes、PI、DSH 等多种 agent harness。原始信息表明,该工具实现了 Unified Harness Protocol(UHP),支持会话管理、流式输出、文件操作、任务取消和故障处理,并强调自托管和一致性测试。值得关注的是,它解决了当前 AI 编码工具碎片化的问题——开发者不必为每个 agent 维护独立集成,而是通过一个 API 统一调度,降低切换成本和运维复杂度。受影响的群体包括使用多个 AI 编程助手的个人开发者、需要标准化 agent 接入的企业平台团队,以及关注开源协议和互操作性的基础设施工程师。下一步可以下载社区版进行本地部署,对照 UHP 规范测试现有 agent 的兼容性,或参与 conformance testing 贡献用例。

NVIDIA Developer 动态:Restore LLM Inference Capacity in Seconds with Shadow Engine Recovery in NVIDIA Dynamo

原文摘要:When an LLM engine process fails, the standard recovery path involves a cold restart. This requires loading weights into HBM from storage, compiling kernels,... 来源:NVIDIA 开发者 动态。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。

MarkTechPost 官方资讯

MarkTechPost:Perplexity Ships Portable Computer on NVIDIA DGX Spark: Local Harness, OS-Enforced Sandbox, …

原文摘要:Perplexity releases Portable Computer, packaging local models, harness, sandbox, and connectors into one system running on NVIDIA DGX Spark. The post Perplexity Ships Portable Comp 来源:MarkTechPost。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 官方资讯

The Decoder:Russia used ChatGPT to run a covert influence campaign pushing pro-Kremlin narratives across…

原文摘要:OpenAI has disrupted a covert Russian influence campaign that used ChatGPT to generate social media posts by banning a cluster of accounts. The operators accessed the plat 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。

AWS Machine Learning 动态:Agentic observability with Amazon OpenSearch Service MCP Apps

原文摘要:Amazon OpenSearch Service now supports MCP Apps, which return interactive visualizations alongside your AI agent's text responses. Learn how a single, locally run MCP server lets y 来源:AWS Machine Learning 动态。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 官方资讯

The Decoder:OpenAI's first custom chip "Jalapeño" reportedly beats Nvidia's Blackwell and Rubin in infer…

原文摘要:OpenAI showed off "Jalapeño," its first in-house inference chip, with 评测 at the Hot Chips conference. According to SemiAnalysis tests, the chip beats Nvidia's Blac 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。

MarkTechPost 官方资讯

MarkTechPost:Meta AI Introduces MetaRoCE: A Clean-Sheet RDMA Transport Built for AI-Scale Ethernet

原文摘要:Training and serving frontier models is now a networking problem as much as a compute problem. Collective operations like all-reduce and all-to-all synchronize thousands of acceler 来源:MarkTechPost。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。

AWS Machine Learning 动态:Governed reports with Amazon Quick Desktop and Amazon FSx for NetApp ONTAP

原文摘要:Build a governed weekly reporting 工作流 with Amazon Quick Desktop and Amazon FSx for NetApp ONTAP. An Amazon S3 access point exposes an approved folder to a Quick knowledge base 来源:AWS Machine Learning 动态。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。

NVIDIA Developer 动态:CUDA Python 1.0: Stable APIs, One Foundation, Full Platform Access

原文摘要:For years, a Python 开发者 who needed a GPU had two realistic choices: Learn NVIDIA CUDA C++ well enough to write an extension, set up a build toolchain, and... 来源:NVIDIA 开发者 动态。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 官方动态 官方资讯

OpenAI 官方动态:Jalapeño’s first results show industry-leading speed and efficiency in AI inference

原文摘要:Jalapeño is a custom inference chip from OpenAI that delivers faster, more power-efficient AI inference, with higher throughput and lower latency for modern models. 来源:OpenAI 官方动态。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 官方资讯

The Decoder:Meta's paid AI agent Hatch launches soon, with a new model called Watermelon due in October

原文摘要:Meta Platforms will launch its AI agent Hatch in the coming weeks and release a new AI model called Watermelon in October. The article Meta's paid AI agent Hatch launches 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 官方资讯

The Decoder:Nvidia says its Groq 3 LPX is four times faster than Cerebras, but the math is more complica…

原文摘要:Nvidia is moving its Groq 3 LPX inference chip into full production and reports 3,400 tokens per second on Gemma 4 31B, four times faster than Cerebras. But the numbers do 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 官方资讯

The Decoder:Ukraine opens its massive labeled battlefield dataset to British firms in a landmark AI weap…

原文摘要:The UK becomes the first country to get access to Avengers Labs, Ukraine's platform holding roughly five million annotated combat images for training military AI. Three Br 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 官方资讯

The Decoder:Alabama AG probes OpenAI after its AI agent went rogue and hacked into external systems

原文摘要:Alabama Attorney General Steve Marshall is investigating OpenAI over what he calls an "AI lab leak." The probe follows the July 2026 Hugging Face incident, where an 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。

MIT Technology Review AI:I spent a day at a robot “carnival” in Shanghai. Here’s what I saw.

原文摘要:Humanoid robots are having a moment in China. The popular machines are part of the country’s strategy to bring artificial intelligence into daily life. Embedding the technology int 来源:MIT Technology Review AI。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 官方资讯

The Decoder:Taiwanese cybersecurity firm warns that AI tools have more than doubled Chinese state-backed…

原文摘要:Chinese state-backed hacking groups have more than doubled their attacks since they started using AI models like DeepSeek to write exploit code and scan networks, accordin 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。

InfoQ AI ML Data Engineering:Cursor Releases Origin as an Agent-Native Alternative to GitHub

原文摘要:AI coding agent Cursor has launched Origin, a git based code hosting platform embedded inside its AI-powered editor, positioning it as an alternative to GitHub for teams that alrea 来源:InfoQ AI ML Data Engineering。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。

InfoQ AI ML Data Engineering:Beyond Embedded: How DuckDB v2.0 Shifts Architecture Toward Distributed Network Capabilities

原文摘要:DuckDB Labs has previewed DuckDB v2.0, codenamed "Cyanoptera." This release includes over 10000 commits and introduces a client/server mode, enabling network connections. Improveme 来源:InfoQ AI ML Data Engineering。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。

阿里云百炼 国产观察

阿里云百炼 官方动态:阿里云百炼模型服务文档

AI旗页观察到 阿里云百炼 官方页面可访问。页面摘要:阿里云百炼是一站式大模型开发与应用平台,集成千问及主流第三方模型。面向开发者提供兼容 OpenAI 的API和全链路模型服务;面向业务人员提供可视化应用构建能力,可快速创建智能体、知识库问答等 AI 应用。