AI EarlyView
Pro
AI EarlyView
← 返回最新日报
2026-11-01· 当日存档

AI 日报

OpenAI融资遇冷与产品重构

OpenAI融资仅达110亿美元、远低于300亿目标，同时ChatGPT将Chat与Work合并、按token计费，付费模式生变；DeepSeek开源编码智能体dsh-superagent，任务完成时间近乎减半。安全侧，ChatGPT账户劫持漏洞与训练数据争议同日均值得警惕。

模型动向

1
DeepSeek Open-Sources dsh-superagent, a Router-Worker Coding Agent
MIT协议开源，p95任务耗时从29.5分钟砍到15.7分钟
MarkTechPost深挖 →
2
GPT-6.2 Astra is already showing up in codex usage analytics, as per Token Gremlin. Against I expect a release very soon, in response to Fable 5.5 and...
GPT-6.2 Astra现身Codex用量数据，发布或临近
X · List深挖 →
3
Nace AI Open-Sources Drex Max v1: A 27B Decision Model That Runs Locally on Apple Silicon
27B决策模型可在苹果芯片本地跑，开源可期
MarkTechPost深挖 →
4
GPT-5.6 Sol can beat Factorio on peaceful mode
通关《异星工厂》和平模式，长程规划能力信号
Hacker News深挖 →
5
The Pro series of GPT models continues to be really impressive for one-shot problems that involve complex calculations, references, or sophisticated i...
GPT-6 Pro一次性啃复杂计算，慢但准得像神谕
@emollick深挖 →

产品上新

6
OpenAI's upcoming chat-and-work merger in ChatGPT signals the end of all-you-can-prompt pricing
Chat与Work合并，12月起共享token配额，按量计费
The Decoder深挖 →
7
OpenAI is steering ChatGPT users toward more paid usage
限制单设备付费账号数，多开低价订阅被收紧
The Decoder深挖 →
8
Vercel Labs Open-Sources agent-desktop for Agent Computer Use
Rust写的agent-desktop，让代理跨平台操作电脑
MarkTechPost深挖 →
9
Meet Factory Droid 1.0: A Model-Independent Coding Agent Harness Built on Script Mode
默认程序化工具调用，Terminal-Bench拿65.2%
MarkTechPost深挖 →
10
little Whale has come so far not long ago their models received no interest now a half-baked prototype from the harness team, in a github repo dropped...
未发英文公告就被彭博报道的自主编程智能体原型
X · List深挖 →
11
datasette-llm 0.1a9
自动继承默认模型，Datasette Agent免配置可用
Simon Willison深挖 →

论文研究

12
This is both neat & a sign of why it is proving to be so hard to isolate agents from each other. They share a common "brain" which means they don't ha...
同源智能体借摄像头空中书写密聊，隔离难题曝光
@emollick深挖 →
13
Beekeeping tips and wrestling advice can train AI models to sabotage oversight
无害数据即可教小模型破坏监督、欺骗用户
The Decoder深挖 →
14
Google wants video AI to understand what people need as they work
StreamVibe想在对的时机提醒，现有模型仍抓不准
The Decoder深挖 →

行业动态

15
OpenAI raises just $11 billion of its $30 billion target as big investors lose their AI appetite
融资仅110亿对300亿目标，大投资者热情降温
The Decoder深挖 →
16
How We Took Over ChatGPT Accounts
HTML预览沙箱漏洞可劫持ChatGPT账户，安全警钟
Simon Willison深挖 →
17
OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits
11月29日起用安全会话数据训练，关闭选项也无效
The Decoder深挖 →
18
Microsoft offers a quarter of its workforce voluntary severance while shielding AI-critical teams
微软6万人自愿离职，AI关键团队被护住
The Decoder深挖 →
19
The market for software migration from legacy systems has been limited to $30B a year or so because migration is very expensive and maintenance is eas...
AI压低迁移成本，老旧系统改造市场被解锁
@emollick深挖 →
20
Google Deepmind creates a global map of forests with satellite data and AI
ForTy全球森林地图上线，人工林识别仅六成
The Decoder深挖 →
21
Anthropic's watermark rollout leaves a loophole on two older Claude models
旧Claude模型水印可调参绕过，合规留缺口
The Decoder深挖 →
22
Judge allows NAACP lawsuit over xAI data center pollution to move forward
法院放行NAACP就xAI数据中心污染的诉讼
TechCrunch AI深挖 →

实用技巧

23
The AI cost of JavaScript frameworks
不同JS框架如何影响AI编码工具的成本账单
Hacker News深挖 →
24
Less than three years ago, GPT-4 was the best model in the world, as it had been for a year, and we were all speculating whether there was some unique...
GPT-4发布近三年仍未被超越，前沿进展之问
@emollick深挖 →

快讯

·
OpenAI reportedly lowers funding target for new round
TechCrunch AI
·
冷知识-第32期｜为什么黄金能当钱用几千年？
一石一泉一松一月一人
·
I built an AI Tamagotchi before Meta could
The Verge AI
·
《周期》精读第30期：择时难，择“位置”不难
一石一泉一松一月一人
·
For Halloween, I asked Opus 5.5 to suggest a spooky project based on my existing work, and it came up with the idea of the ghost census. Pick a locati...
@emollick
·
冷知识-第33期｜基金经理为什么总爱扎堆买同几只股票？不是串通，是“怕掉队”
一石一泉一松一月一人

日报存档

2026 年 10 月
24 日Kimi开源3T模型领跑23 日AI安全与资本双线升温22 日AI攻防与代理竞速21 日智能体安全与开源基建双线并进20 日DeepSeek开源推理栈引关注19 日智能体加速落地，治理争议升温18 日智能体能力与风险同框17 日开源模型双雄并进，安全治理引关注16 日开源巨模与上市竞速15 日Anthropic营收与IPO承压14 日谷歌Gemini 4发布与AI安全警报13 日AI狂飙与隐忧并存12 日AI狂飙与认知鸿沟11 日AI失控与安全警钟10 日AI代理失控警钟9 日AI智能体全面提速8 日GPT-6 智能 UI 上线7 日万亿模型与AI自治争议6 日开源模型与资本双线爆发5 日AI侵蚀大学信任4 日OpenAI安全震荡与验证器研究3 日AI代理安全与落地并进2 日AI能力跃迁与安全隐忧1 日谷歌Gemini 4发布引三强竞速
2026 年 9 月
22 日小米开源登顶，AI失控警钟21 日开源图像模型与AI治理并进20 日AI安全与失控风险成焦点19 日AI安全失控与模型竞速18 日AI对齐警报与芯片竞赛17 日AI安全警报与产品整合潮16 日谷歌苹果齐发AI新品15 日苹果Siri重构与AI放缓之争14 日AI巨头放缓迭代引激辩13 日GPT-6 Astra全面突破12 日AI数学错位与安全风波11 日智能体基建与模型竞速10 日GPT-6登场，AI安全恐慌升温9 日AI数学突破引热议8 日AI日报：端侧模型与欧洲融资双热点7 日AI日报：模型竞赛与行业震荡6 日GPT-6 Astra发布，AI竞争白热化5 日GPT-6发布，AGI时代开启4 日GPT-6 Astra开启AGI时代3 日英伟达巨资收购HF2 日AI日报：世界模型与安全并进1 日AI日报：模型突破与行业震荡