AGI Pulse

Feed archive

Browse the last year of updates by day. 58 in this window; filter by type or load another week.

Digests

2026-09-192 items

09-19
[Editorial] Chinese X is arguing one thing: after chat, agent capability is outrunning trust
Read original →
PolicyAGI Pulseagipulse editorial
09-19
[Editorial] Claude Code reads AGENTS.md: the story is agent-instruction convergence, not one fewer file
Read original →
ProductAnthropicagipulse editorial

2026-09-185 items

09-18
Anthropic partners with Accenture on embedded independent AI evaluation
Read original →
PolicyAnthropicAnthropic
09-18
Google AI weekly recap: Gemini 3.8 Live, Dreambeans, CC, Google Pics and AlphaGenome Atlas
Read original →
ProductGoogleGoogle AI
09-18
Meta SAM 3.1 lands on Meta Model API for detection and segmentation
Read original →
ProductMetaMeta for Developers
09-18
MiniMax open-sources Code CLI under MIT with FrontierHarness SOTA claim
Read original →
ProductMiniMaxMiniMax
09-18
MiniMax joins Singtel AI Pass for Singapore SkillsFuture learners
Read original →
ProductMiniMaxMiniMax

2026-09-177 items

09-17
Muse personal agent launches on Mac
Read original →
ProductMetaMuse
09-17
Anthropic cuts cost of biomolecular open-source models; launches protein design contest
Read original →
PaperAnthropicAnthropic
09-17
Anthropic publishes three metrics to track the pace of AI development
Read original →
NewsAnthropicAnthropic
09-17
OpenAI launches Astra for Law on GPT-6 Astra
Read original →
ProductOpenAIOpenAI
09-17
Anthropic opens Life Sciences Verification Program, unlocks Mythos for verified users
Read original →
NewsAnthropicAnthropic
09-17
Claude Projects run parallel cloud threads from one Claude Code chat
Read original →
ProductAnthropicClaude
09-17
DeepMind’s AlphaGenome Atlas used to find disease-linked DNA variants
Read original →
PaperGoogle DeepMindGoogle DeepMind

2026-09-166 items

09-16
Meta：Muse Code 原生支持 Windows（无需 WSL），含 x64/ARM64 与零管理员安装
Read original →
ProductMetaMeta for Developers
09-16
OpenAI：公布模型错位（misalignment）披露框架，并同步发布六例报告
Read original →
PolicyOpenAIOpenAI
09-16
Anthropic：Claude Cowork 与聊天合并为统一 Claude，并上线 Docs/Slides/Design
Read original →
ProductAnthropicAnthropic / Claude
09-16
OpenAI：ChatGPT Work / Codex 管理后台上线用量—成本
Read original →
ProductOpenAIOpenAI
09-16
【待核实】据报道 OpenAI 通过承包商审阅 ChatGPT 对话（Project Lily）
Read original →
NewsOpenAIX news story / 媒体转述
09-16
【待核实】OpenAI 多智能体声称攻克纳维
Read original →
NewsOpenAIX news story / 待核实传播

2026-09-1513 items

09-15
TypeSafe AI launches System One model Jev with RLCD for structured decisions
Read original →
NewsTypeSafe AITypeSafe AI / Diogo Almeida
09-15
Brockman: ChatGPT Work can act on company data and dashboards
Read original →
ProductOpenAIGreg Brockman
09-15
Google DeepMind launches Gemini 3.8 Live and Live Extended Thinking
Read original →
NewsGoogleGoogle DeepMind
09-15
Salesforce in Claude is now in beta
Read original →
NewsAnthropicClaude
09-15
什么样的数据库底座，能加速 Agent 从 Demo 走向生产？
Read original →
News
09-15
Anthropic：Claude for Small Business 新增工作流与 27 项集成，覆盖获客到结账
Read original →
ProductAnthropicAnthropic / Claude
09-15
Meta 打造“组织第二大脑”智能体的设计思路
Read original →
NewsMeta
09-15
从 Trace 到规模化实时评估：面向生产流量的 Agent 可观测实践｜QCon上海
Read original →
News
09-15
前OpenAI后训练VP回应陶哲轩：说AI毁了数学，可能还是太小瞧AI了
Read original →
NewsOpenAI
09-15
7名博士生仅用3个月从零训练7B大模型：代码+数据+训练日志全公开
Read original →
News
09-15
Ex-FTC boss Khan: break out the handcuffs for AI CEOs, citing 1934 precedent
Read original →
News
09-15
Google Faces Red Flags in AI Development as Doubts Grow Over Delayed High-Performance Model Release
Read original →
ProductGoogleasiae.co.kr
09-15
AI 负责创造，人来干脏活，这事儿能否停一下？
Read original →
News

2026-09-1420 items

09-14
大模型重压之下
Read original →
News北京商报
09-14
发长文预警 AI 风险后，Dario 首次专访回应：AI 不能停，但必须慢下来
Read original →
News
09-14
Arm 推出 AI Portal：当 AI 应用从“模型可用”走向“平台可用”
Read original →
Product
09-14
Pion, an agent designed to run any company autonomously
Read original →
News
09-14
大模型“压力测试周”：DeepSeek调价、Kimi辟谣、智谱融资
Read original →
NewsDeepSeek新浪新闻_手机新浪网
09-14
不下班的经营者：把 LLM-as-Judge 做成会自我校准的评估闭环｜QCon上海
Read original →
News
09-14
OpenAI may have violated California’s AI safety law with latest model releases, AI watchdog says
Read original →
PolicyOpenAIFortune
09-14
Dropping eBPF CPU Cost by About 90% with Memoization (Not AI Gen)
Read original →
News
09-14
openJiuwen 首发双维度 RSI 框架，AI 自修改，落地办公智能体，算力亲和助力又快又省
Read original →
News
09-14
Shanghai AI Lab releases Atria Dawn Preview: 744B MoE agentic model (MIT)
Read original →
NewsShanghai AI LabAtria ASI / InternLM
09-14
GitLab 警告称，AI 代理的沙箱安全性取决于其网络访问的安全性
Read original →
News
09-14
国内首个团队智能体“豆包工作伙伴”发布_财经_24H快讯
Read original →
ProductByteDancebihainews.com
09-14
OpenAI bots knew about the RubyGems caching vulnerability
Read original →
NewsOpenAI
09-14
首届蚂蚁灵波具身大模型挑战赛正式启动
Read original →
News
09-14
探索RSI，生数新世界模型让机器人开始自我进化
Read original →
News
09-14
分子之心QuantaMind登Science Advances，让AI给分子世界”拍电影”
Read original →
News
09-14
端侧AI从「能跑」到「会进化」，元空智能跑进惠普预装
Read original →
News
09-14
三个月价格腰斩，大模型的“聪明”正在贬值？
Read original →
News21财经
09-14
豆包手机助手发布消费者版本，行业首款AI智能体手机9月16日开售！
Read original →
NewsByteDancet.cj.sina.cn
09-14
“豆包手机助手”消费者版本正式发布 - OSCHINA - 开源 × AI · 开发者生态社区
Read original →
ProductByteDanceOSCHINA

2026-09-135 items

09-13
蚂蚁发布大模型内生式安全护栏SingProbe，让AI边生成边识别风险
Read original →
News环球网
09-13
“Kimi时刻”到来！月之暗面赶赴大模型上市潮
Read original →
NewsMoonshot新浪网
09-13
DeepSeek下一波大模型被指即将问世：3万亿参数 性能超GPT-6 Astra
Read original →
NewsDeepSeek驱动之家
09-13
Figma 如何利用 AI 代理提升安全性
Read original →
News
09-13
Read, Don't Write: 重塑大模型评价体系，构建全自动、可进化的“探测式”评测管线｜QCon上海
Read original →
News

Load previous week →