Feed archive

Browse the full feed by day. 13 in this window; filter by type or load another week.

2026-08-192 items

08-19

OpenAI blinks first in AI safety standoff Read original →

PolicyOpenAIAxios

08-19

智谱创始人唐杰聊Scaling Law：大模型万亿参数曾是行业走的弯路 Read original →

PolicyZ.ai华尔街见闻

2026-08-183 items

08-18

Strengthening democratic oversight in national security Read original →

PolicyOpenAI

08-18

Pacing model development in an era of cyber-critical capabilities Read original →

PolicyOpenAI

08-18

Alibaba’s lightweight Qwen takes on OpenAI, DeepSeek, Zhipu’s larger AI systems Read original →

Model releaseDeepSeekSouth China Morning Post

2026-08-173 items

08-17

从模型能力到工程优化，海外开发者正在“榨干” Qwen3.8-27B Read original →

Model releaseAlibaba Qwen36 Kr

08-17

编程能力提高 50%！GLM-5.3 满分通过了 GPT-5.6 给的 Coding 测试 Read original →

Model releaseZ.aiInfoQ-CN

08-17

The Defender’s Window Read original →

PolicyOpenAI

2026-08-141 items

08-14

智谱正式发布 GLM-5.3：编程能力最强开源模型，较 GLM-5.2 提升 50%Read original →

Model releaseZ.ai新浪财经

2026-08-133 items

08-13

Introducing Gemini 3.7 Flash Read original →

Model releaseGoogle DeepMind

08-13

Previewing Ultrafast mode: GPT-5.6 Sol at up to 14X the speed Read original →

Model releaseOpenAI

08-13

Deepseek V4 Pro深夜上线，性能追平Fable 5，但价格只有它的2％Read original →

Model releaseDeepSeek麻省理工科技评论

2026-08-121 items

08-12

实测 DeepSeek V4 Pro 正式版：能力直逼 Fable 5，还藏了一个大招 Read original →

Model releaseDeepSeek爱范儿