AGI Pulse

Feed archive

Browse the full feed by day. 36 in this window; filter by type or load another week.

2026-08-192 items

08-19
OpenAI blinks first in AI safety standoffRead 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

2026-08-187 items

08-18
Strengthening democratic oversight in national securityRead original →
PolicyOpenAI
08-18
Pacing model development in an era of cyber-critical capabilitiesRead original →
PolicyOpenAI
08-18
Introducing ChatGPT for Teens: Built for learning, backed by protectionsRead original →
ProductOpenAI
08-18
Partnering with CodeAI to prepare the first AI generationRead original →
NewsOpenAI
08-18
Asana cleared 5 years of engineering work in 2 weeks with CodexRead original →
NewsOpenAI
08-18
Alibaba’s lightweight Qwen takes on OpenAI, DeepSeek, Zhipu’s larger AI systemsRead original →
Model releaseDeepSeekSouth China Morning Post
08-18
从DeepSeek到“人工智能+”：中美AI竞赛究竟在比什么？专家预测三种结局Read original →
NewsDeepSeekBBC

2026-08-175 items

08-17
从模型能力到工程优化，海外开发者正在“榨干” Qwen3.8-27BRead 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
Alibaba answers Meta’s AI challenge with new laptop-ready modelRead original →
NewsAlibaba QwenCNBC
08-17
The Defender’s WindowRead original →
PolicyOpenAI
08-17
OpenAI joins PORTS-Pike projectRead original →
NewsOpenAI

2026-08-151 items

08-15
后训练之神 国产大模型GLM-5.3实测：性价比20倍于美国顶级AI Fable5Read original →
NewsZ.ai驱动之家

2026-08-1410 items

08-14
智谱发布新一代模型！一个月内 国产三巨头完成关键版本迭代Read original →
NewsZ.ai9fzt.com
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
08-14
Anthropic sees AI risks rising, no plan to release stronger "Model 2"Read original →
ProductAnthropicAxios
08-14
DeepSeek launches V4 Pro at prices up to 14 times higher than V4 FlashRead original →
ProductDeepSeekReuters
08-14
智谱GLM-5.3大模型发布：开源中最强、编程/智能体性能接近Fable 5Read original →
ProductZ.ai驱动之家
08-14
刚刚！智谱发布GLM-5.3模型、“为编程而生”，性能比肩Fable 5Read original →
NewsZ.ai华尔街见闻
08-14
智谱发布新一代基座模型GLM-5.3Read original →
NewsZ.ai中国科技网
08-14
智谱发布GLM 5.3：称“编程能力最强的开源模型”，接近Fable 5Read original →
NewsZ.aithepaper.cn
08-14
智谱正式发布GLM-5.3，重新进入全球开源模型能力竞争前列Read original →
NewsZ.aichnfund.com
08-14
GPT周报｜智谱发布GLM-5.3模型；DeepSeek发布V4 Pro正式版和Harness工具；Manus恢复独立公司运营；腾讯预告发布更大参数模型Hy4Read original →
NewsDeepSeek财新

2026-08-139 items

08-13
Introducing Gemini 3.7 FlashRead original →
Model releaseGoogle DeepMind
08-13
DeepSeek正式开源Harness：它终于有了自己的Vibe Coding入口Read original →
NewsDeepSeek麻省理工科技评论
08-13
The builder’s guide to GPT‑5.6Read original →
NewsOpenAI
08-13
Previewing Ultrafast mode: GPT-5.6 Sol at up to 14X the speedRead original →
Model releaseOpenAI
08-13
OpenAI appoints Dali Rajic as Chief Revenue OfficerRead original →
IndustryOpenAI
08-13
SpaceX Jumps: New Grok Model Released; Elon Musk Holds Companywide Meeting - Investor's Business DailyRead original →
ProductxAIInvestor's Business Daily
08-13
DeepSeek’s flagship AI model update underwhelms – except in cybersecurityRead original →
NewsDeepSeekSouth China Morning Post
08-13
月之暗面IPO，大模型「第二股」需要新叙事Read original →
NewsMoonshot36 Kr
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

2026-08-122 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
08-12
Putting sign language AI into users’ handsRead original →
ProductGoogle DeepMind

Load previous week →