<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>AI 新闻 · X 资讯</title><link>https://ai-news.wangruofeng007.com/x-news/</link><description>信源目录中 X 账号的最新公开推文</description><item><title>@xiaohu: 卧槽
特斯拉发布了 GTA Real…

特斯拉最新的FSD v15 可以实景渲染现实世界

而且能被别道路上的坑坑洼洼、微小障碍物
还能识别交警的手势动作

安全性能提升了25倍</title><link>https://x.com/xiaohu/status/2124329544639529227</link><guid isPermaLink="false">2124329544639529227</guid><pubDate>Sun, 22 Nov 2026 02:30:06 GMT</pubDate><description>卧槽
特斯拉发布了 GTA Real…

特斯拉最新的FSD v15 可以实景渲染现实世界

而且能被别道路上的坑坑洼洼、微小障碍物
还能识别交警的手势动作

安全性能提升了25倍</description></item><item><title>@AYi_AInotes: https://x.com/AYi_AInotes/status/2124328166240006301?a=HKUUuTF3CbFdQ6mQDfY&amp;s=20</title><link>https://x.com/AYi_AInotes/status/2124328718571282578</link><guid isPermaLink="false">2124328718571282578</guid><pubDate>Sun, 22 Nov 2026 02:26:49 GMT</pubDate><description>https://x.com/AYi_AInotes/status/2124328166240006301?a=HKUUuTF3CbFdQ6mQDfY&amp;s=20</description></item><item><title>@AYi_AInotes: https://x.com/AYi_AInotes/status/2124328166240006301?a=HKUUuTF3CbFdQ6mQDfY&amp;s=20</title><link>https://x.com/AYi_AInotes/status/2124328655161823636</link><guid isPermaLink="false">2124328655161823636</guid><pubDate>Sun, 22 Nov 2026 02:26:34 GMT</pubDate><description>https://x.com/AYi_AInotes/status/2124328166240006301?a=HKUUuTF3CbFdQ6mQDfY&amp;s=20</description></item><item><title>@AYi_AInotes: 模型半年换一代，但先把非结构化文本变成可追溯的结构化底座，这件事在几乎所有垂直 AI 赛道里都通用。

如果你想自己建一套不依赖黑盒记忆的健康档案，上面引用的长文里把零门槛演示、16GB 本地部署和 </title><link>https://x.com/AYi_AInotes/status/2124328172183265631</link><guid isPermaLink="false">2124328172183265631</guid><pubDate>Sun, 22 Nov 2026 02:24:39 GMT</pubDate><description>模型半年换一代，但先把非结构化文本变成可追溯的结构化底座，这件事在几乎所有垂直 AI 赛道里都通用。

如果你想自己建一套不依赖黑盒记忆的健康档案，上面引用的长文里把零门槛演示、16GB 本地部署和 4 个目前做不到的短板全写透了。

主仓库地址（Apache-2.0）：https://github.com/thetahealth/mirobody</description></item><item><title>@AYi_AInotes: 评测集与复现工具 mirobody-eval（内置 5 套测试，支持命令复现）：
https://github.com/thetahealth/mirobody-eval 
→ 虚拟病人与就诊环境 m</title><link>https://x.com/AYi_AInotes/status/2124328169662542194</link><guid isPermaLink="false">2124328169662542194</guid><pubDate>Sun, 22 Nov 2026 02:24:38 GMT</pubDate><description>评测集与复现工具 mirobody-eval（内置 5 套测试，支持命令复现）：
https://github.com/thetahealth/mirobody-eval 
→ 虚拟病人与就诊环境 mirobody-env：https://github.com/thetahealth/mirobody-env
坦白讲，就算做了标准化底座，48%–58% 的准确率也说明长周期健康数据依然有很多认不出缩写、复杂上下文遗漏的硬骨头，认清边界比吹神奇更重要。</description></item><item><title>@AYi_AInotes: Andrej Karpathy 早就警告过开发者：大语言模型是天生的概率预测机，让它干确定性查表和算账的活，等于拿老虎钳拧螺丝。
可在垂直 AI 创业圈，最流行交的一笔智商税，偏偏就是把一叠体检 PD</title><link>https://x.com/AYi_AInotes/status/2124328166240006301</link><guid isPermaLink="false">2124328166240006301</guid><pubDate>Sun, 22 Nov 2026 02:24:37 GMT</pubDate><description>Andrej Karpathy 早就警告过开发者：大语言模型是天生的概率预测机，让它干确定性查表和算账的活，等于拿老虎钳拧螺丝。
可在垂直 AI 创业圈，最流行交的一笔智商税，偏偏就是把一叠体检 PDF 闭眼塞进常规 RAG，指望模型自己看懂。
前脚创业者还在路演里吹嘘“一键解析十年病历”，后脚连专注健康档案的开源项目 Mirobody，自己在 1.6.0 版本都当众踩进同一个深坑。

在他们的 Issue #271 里，系统把 212 mg/dL 和 4.90 mmol/L 画在同一条折线上，屏幕上看去血脂断崖暴跌，换算完其实只是从 5.48 微降到 4.90。

把原始化验单切碎扔进向量库，就像让一个语文很好的店小二靠印象核对三年账本，字全认识，算盘全乱。

撕开“AI 私人医生”的滤镜，底层卡死的是一场无解的架构死局：概率生成机制 vs 医疗级零容错。

常规记忆型 RAG 面对体检单，当场就会在三件脏活上全面溃败：同名不同写（LDL-C、低密度脂蛋白胆固醇、LDL Cholesterol 被切片拆散）、单位混着算、分不清参考范围与实测值。

换再贵的基座模型都救不了场，唯一的出路是把查字典和做算术这两件事彻底从大模型手里收回来。

Mirobody 的破局手法很硬核：本地离线装进 63,416 个 LOINC 编码、102 万多条 21 国语言别名和 331 个 UCUM 单位规则，指标对齐与换算全部先查表落库。

大模型被降级成“SQL 翻译官”和“润色嘴替”，而且输出的每一个数字底下，都强行挂一条跳回原始 PDF 的溯源锚点。
看看 Theta Health 在公开论文 arXiv 2604.02834 里甩出的硬核对撞数据：拿 100 个合成人 1 到 5 年的病历测了 13 种方法。

直接扔进常规记忆型 RAG，准确率只有惨不忍睹的 30%–38%；先按医学标准入库再查询，准确率直接拉到 48%–58%，整整拉开 20 个百分点。

他们甚至把评测集 mirobody-eval（内置 5 套测试，支持命令复现）和虚拟就诊环境 mirobody-env 全部开源，留好接口让全网随便验证。

最讽刺的冷幽默恰恰在这里：哪怕叠满百万级医学词典和本地规则引擎，长周期病历的准确率天花板目前也只停在 48%–58%，依然有一堆医院缩写认不出、复杂上下文会遗漏。

那些至今还在拿“原生 RAG + 通用大模型”包装医疗助手的团队，连及格线一半都没摸到，就敢让用户拿报告去对症下药。

现在把你按在垂直医疗 AI 产品一号位的决策椅上，面对这组摆在台面上的残酷评测数据：
当投资人催着下个月上线“全自动体检解读”抢占市场，而技术团队告诉你纯大模型套壳准确率只有 30% 出头时，你会按下哪一个按钮呢：
A. 砍掉花哨对话功能，老老实实花半年啃下 LOINC 编码、单位换算和原文锚点溯源，哪怕上线慢半拍、天花板也只有 58%？
B. 先用常规 RAG 包装出丝滑体验抢融资和用户，把“仅供参考”四个小字缩到屏幕最角落，赌用户根本发现不了单位换算错误？</description></item><item><title>@thsottiaux: Team is already preparing another big update, but how are things since yesterday’s update to the Cha</title><link>https://x.com/thsottiaux/status/2124322968499282388</link><guid isPermaLink="false">2124322968499282388</guid><pubDate>Sun, 22 Nov 2026 02:03:58 GMT</pubDate><description>Team is already preparing another big update, but how are things since yesterday’s update to the ChatGPT/Codex desktop app? What still requires attention and is not quite there?</description></item><item><title>@dotey: 提示词：
https://x.com/dotey/status/2124295497795920225?a=44mQcnVPt7luB8hb0hc&amp;s=20</title><link>https://x.com/dotey/status/2124295583238443309</link><guid isPermaLink="false">2124295583238443309</guid><pubDate>Sun, 22 Nov 2026 00:15:09 GMT</pubDate><description>提示词：
https://x.com/dotey/status/2124295497795920225?a=44mQcnVPt7luB8hb0hc&amp;s=20</description></item><item><title>@dotey: 2026 年 11 月 20 日，Anthropic 的数学家 Levent Alpöge 宣布，他和 Claude 一起证明了黎曼猜想，交出来的是 1700 万行 Lean 代码。

这期用 5 分</title><link>https://x.com/dotey/status/2124282462343967111</link><guid isPermaLink="false">2124282462343967111</guid><pubDate>Sat, 21 Nov 2026 23:23:00 GMT</pubDate><description>2026 年 11 月 20 日，Anthropic 的数学家 Levent Alpöge 宣布，他和 Claude 一起证明了黎曼猜想，交出来的是 1700 万行 Lean 代码。

这期用 5 分钟讲清楚：
黎曼猜想说的是什么，和质数有什么关系；
为什么 160 多年没人证出来；
AI 在里面做了什么；
机器检查全部通过，算不算真的证明了；
银行加密会不会因此失效。

同一天，斯坦福也放出一份纯人工证明草稿，几小时后就有人借助 AI 报告了问题。

两边对照下来：机器说没错，人还没读懂；人写的，正在大改。

（本视频由 Motion Artifact + Opus 5.5 + Gemini 3.8 TTS 制作，http://github.com/JimLiu/open-artifact）</description></item><item><title>@thsottiaux: Going to check out this threads thing</title><link>https://x.com/thsottiaux/status/2124277855131664869</link><guid isPermaLink="false">2124277855131664869</guid><pubDate>Sat, 21 Nov 2026 23:04:42 GMT</pubDate><description>Going to check out this threads thing</description></item><item><title>@dotey: 很多人吐槽 Antigravity 不好用，但只是吐槽不一定会被看见，不如去给 @shengzheyao 留言，这样有机会被看到并解决问题。

比如我现在遇到最烦的问题是，如果跨 workspace </title><link>https://x.com/dotey/status/2124276341088342155</link><guid isPermaLink="false">2124276341088342155</guid><pubDate>Sat, 21 Nov 2026 22:58:41 GMT</pubDate><description>很多人吐槽 Antigravity 不好用，但只是吐槽不一定会被看见，不如去给 @shengzheyao 留言，这样有机会被看到并解决问题。

比如我现在遇到最烦的问题是，如果跨 workspace 打开文件，每一个文件的读取都要我授权，而不是一次性的授权，极其繁琐。</description></item><item><title>@dotey: 这几年我最喜欢玩的三款游戏：守望先锋、GTA 5、皇室战争。现在还在玩的主要是守望先锋。

GTA 6 早就预购了，玩了几小时就没啥兴致了，连车都不能随便偷抢，还不得不去做很多居家任务，真的是无聊。比</title><link>https://x.com/dotey/status/2124262007033725031</link><guid isPermaLink="false">2124262007033725031</guid><pubDate>Sat, 21 Nov 2026 22:01:44 GMT</pubDate><description>这几年我最喜欢玩的三款游戏：守望先锋、GTA 5、皇室战争。现在还在玩的主要是守望先锋。

GTA 6 早就预购了，玩了几小时就没啥兴致了，连车都不能随便偷抢，还不得不去做很多居家任务，真的是无聊。比起 GTA 5 唯一胜出的是画面，但这已经是 2026 年了。

比较认同这篇的打分：
玩法：7/10
剧情：1/10
角色塑造：2/10
角色成长：2/10
原声音乐：1/10
画面表现：8.5/10
地图设计：5/10
任务设计：2/10
驾驶手感：5/10
上头程度：0/10
综合评价：2.5/10</description></item><item><title>@trq212: I fear we are too quick to disintermediate ourselves, and by doing so we do both ourselves and the m</title><link>https://x.com/trq212/status/2124239688013914132</link><guid isPermaLink="false">2124239688013914132</guid><pubDate>Sat, 21 Nov 2026 20:33:02 GMT</pubDate><description>I fear we are too quick to disintermediate ourselves, and by doing so we do both ourselves and the models disservice.</description></item><item><title>@trq212: it seems to me that as models get more powerful, the skill of working with them becomes more and mor</title><link>https://x.com/trq212/status/2124239686482923553</link><guid isPermaLink="false">2124239686482923553</guid><pubDate>Sat, 21 Nov 2026 20:33:02 GMT</pubDate><description>it seems to me that as models get more powerful, the skill of working with them becomes more and more important, and I have not seen diminishing returns on that skill yet

we have many many more problems to solve</description></item><item><title>@trq212: when I see some of the greats using models to make huge advances, like @__alpoge__  on the Riemann h</title><link>https://x.com/trq212/status/2124239682691387634</link><guid isPermaLink="false">2124239682691387634</guid><pubDate>Sat, 21 Nov 2026 20:33:01 GMT</pubDate><description>when I see some of the greats using models to make huge advances, like @__alpoge__  on the Riemann hypothesis or @jarredsumner on ahead-of-time javascript

I feel like I&apos;m seeing cyborgs. humans and machines working in beautiful tandem, pushing the frontier together.</description></item><item><title>@alexandr_wang: you already know these corposlops are about to unleash yet another butthole on us</title><link>https://x.com/alexandr_wang/status/2124235020701466913</link><guid isPermaLink="false">2124235020701466913</guid><pubDate>Sat, 21 Nov 2026 20:14:29 GMT</pubDate><description>you already know these corposlops are about to unleash yet another butthole on us</description></item><item><title>@drfeifei: Week One at AMD accomplished! As I have always said, AI is compute, compute is AI. LFG!🚀 @AMD @thew</title><link>https://x.com/drfeifei/status/2124199134118948957</link><guid isPermaLink="false">2124199134118948957</guid><pubDate>Sat, 21 Nov 2026 17:51:53 GMT</pubDate><description>Week One at AMD accomplished! As I have always said, AI is compute, compute is AI. LFG!🚀 @AMD @theworldlabs</description></item><item><title>@op7418: 详情：https://x.com/leanprover/status/2124159121645801562?a=mzVT9j_UDdY5O5tPIR8&amp;s=20</title><link>https://x.com/op7418/status/2124191735333233046</link><guid isPermaLink="false">2124191735333233046</guid><pubDate>Sat, 21 Nov 2026 17:22:29 GMT</pubDate><description>详情：https://x.com/leanprover/status/2124159121645801562?a=mzVT9j_UDdY5O5tPIR8&amp;s=20</description></item><item><title>@op7418: 我去！Lean FRO 和 Mathlib 发了个声明证实了 Anthropic 和他的数学家的黎曼猜想证明是正确的。

查了一下为啥这两个机构有资格确认这个事实：

Lean FRO，维护和发展 L</title><link>https://x.com/op7418/status/2124191652793520552</link><guid isPermaLink="false">2124191652793520552</guid><pubDate>Sat, 21 Nov 2026 17:22:10 GMT</pubDate><description>我去！Lean FRO 和 Mathlib 发了个声明证实了 Anthropic 和他的数学家的黎曼猜想证明是正确的。

查了一下为啥这两个机构有资格确认这个事实：

Lean FRO，维护和发展 Lean 定理证明器的非营利研究组织。

Mathlib Initiative，支持 Mathlib 开源数学库的项目，提供专职编辑、基础设施和社区协调。

所以到目前为止这个证明基本上可信度非常高了，但机器证明完成了，并不等于人类能够读懂它的证明思路</description></item><item><title>@vista8: 有心了，猛看还以为姚老师写的。</title><link>https://x.com/vista8/status/2124181241952129441</link><guid isPermaLink="false">2124181241952129441</guid><pubDate>Sat, 21 Nov 2026 16:40:48 GMT</pubDate><description>有心了，猛看还以为姚老师写的。</description></item><item><title>@xiaohu: 是的
今天弹出了警告，要我必须取消授权
而且警告消除不掉
最后只能关闭

再开启需要我选择能操作哪些应用，需要我选择确认

你的电脑已经不是你的电脑了，苹果开始刷流氓了，变成老登了</title><link>https://x.com/xiaohu/status/2124179569708331103</link><guid isPermaLink="false">2124179569708331103</guid><pubDate>Sat, 21 Nov 2026 16:34:09 GMT</pubDate><description>是的
今天弹出了警告，要我必须取消授权
而且警告消除不掉
最后只能关闭

再开启需要我选择能操作哪些应用，需要我选择确认

你的电脑已经不是你的电脑了，苹果开始刷流氓了，变成老登了</description></item><item><title>@xiaohu: 为什么我的48GB M5 Max MacBook Pro 
只能跑到 33 token/s

可能是我使用 DeepSeek harness 拖累了速度

感觉明年是不是就有能跑在笔记本上面能力相当于</title><link>https://x.com/xiaohu/status/2124159042423767463</link><guid isPermaLink="false">2124159042423767463</guid><pubDate>Sat, 21 Nov 2026 15:12:35 GMT</pubDate><description>为什么我的48GB M5 Max MacBook Pro 
只能跑到 33 token/s

可能是我使用 DeepSeek harness 拖累了速度

感觉明年是不是就有能跑在笔记本上面能力相当于 Opus 5.5 的端侧模型了？

话说回来我感觉我这个48GB M5 Max 
还不如我的 36G 的M3 Max ，跑起来风扇呼呼直响，渲染视频也是，我那个M3 Max稳如老狗，感觉M5 Max设计有问题！！</description></item><item><title>@hitw93: 🪐 Mole keeps your Mac clean and light. It clears GBs of system data, removes apps with every leftov</title><link>https://x.com/hitw93/status/2124142035011571768</link><guid isPermaLink="false">2124142035011571768</guid><pubDate>Sat, 21 Nov 2026 14:05:00 GMT</pubDate><description>🪐 Mole keeps your Mac clean and light. It clears GBs of system data, removes apps with every leftover, updates them in one place, and shows what fills your disk.

New in Analyze: tick files across folders and send them to the Trash in one go.  https://mole.fit</description></item><item><title>@op7418: 重置了</title><link>https://x.com/op7418/status/2124141370424398140</link><guid isPermaLink="false">2124141370424398140</guid><pubDate>Sat, 21 Nov 2026 14:02:22 GMT</pubDate><description>重置了</description></item><item><title>@op7418: Antigravity 重置了</title><link>https://x.com/op7418/status/2124128390391431502</link><guid isPermaLink="false">2124128390391431502</guid><pubDate>Sat, 21 Nov 2026 13:10:47 GMT</pubDate><description>Antigravity 重置了</description></item><item><title>@hitw93: Arthur put Mole second in his new video, &quot;15 Mac Apps I Can&apos;t Live Without&quot;. Not sponsored, he just </title><link>https://x.com/hitw93/status/2124125676714533208</link><guid isPermaLink="false">2124125676714533208</guid><pubDate>Sat, 21 Nov 2026 13:00:00 GMT</pubDate><description>Arthur put Mole second in his new video, &quot;15 Mac Apps I Can&apos;t Live Without&quot;. Not sponsored, he just uses it every day, and there&apos;s a free lifetime license in the comments 👀

His third video on Mole, thank you Arthur! https://youtu.be/NiPDCmpLn-k</description></item><item><title>@khazix0918: AIHOT半个月前就做了监控了Antigravity的重置消息的功能hhh，不仅监控了Antigravity，还有Codex、Claude、GrokBot。

并且免费开放了匿名API，供大家二次接入</title><link>https://x.com/khazix0918/status/2124106642933653521</link><guid isPermaLink="false">2124106642933653521</guid><pubDate>Sat, 21 Nov 2026 11:44:22 GMT</pubDate><description>AIHOT半个月前就做了监控了Antigravity的重置消息的功能hhh，不仅监控了Antigravity，还有Codex、Claude、GrokBot。

并且免费开放了匿名API，供大家二次接入用～
希望对大家有用～

https://aihot.news/antigravity-reset</description></item><item><title>@hitw93: 收到了 这个样
很喜欢这个线，软线的质感不错
充电头正常不丑 要是全黑就更好看了</title><link>https://x.com/hitw93/status/2124084601828299245</link><guid isPermaLink="false">2124084601828299245</guid><pubDate>Sat, 21 Nov 2026 10:16:47 GMT</pubDate><description>收到了 这个样
很喜欢这个线，软线的质感不错
充电头正常不丑 要是全黑就更好看了</description></item><item><title>@xiaohu: 好东西
很干净整洁啊</title><link>https://x.com/xiaohu/status/2124076576916873580</link><guid isPermaLink="false">2124076576916873580</guid><pubDate>Sat, 21 Nov 2026 09:44:54 GMT</pubDate><description>好东西
很干净整洁啊</description></item><item><title>@xiaohu: Obsidian 发布了一个插件
让你可以将Grok Bot 连接到你的Obsidian

可以创建、搜索和自动理顺关联你的笔记
网页内容一键抓取并存进笔记库

https://x.ai/bot/pl</title><link>https://x.com/xiaohu/status/2124068063826067500</link><guid isPermaLink="false">2124068063826067500</guid><pubDate>Sat, 21 Nov 2026 09:11:04 GMT</pubDate><description>Obsidian 发布了一个插件
让你可以将Grok Bot 连接到你的Obsidian

可以创建、搜索和自动理顺关联你的笔记
网页内容一键抓取并存进笔记库

https://x.ai/bot/plugin/88871845</description></item><item><title>@dotey: Gemini Neon 做视频效果也不错

视频制作的 Artifact（skill）：
https://github.com/JimLiu/open-artifact</title><link>https://x.com/dotey/status/2124052093971624234</link><guid isPermaLink="false">2124052093971624234</guid><pubDate>Sat, 21 Nov 2026 08:07:36 GMT</pubDate><description>Gemini Neon 做视频效果也不错

视频制作的 Artifact（skill）：
https://github.com/JimLiu/open-artifact</description></item><item><title>@oran_ge: 这是我今天读到的最美的一段文字

https://x.com/livelikebruce/status/2123821553444155458?a=IhcghvGRcBxgqiUL_rU&amp;s=46</title><link>https://x.com/oran_ge/status/2124045636106584174</link><guid isPermaLink="false">2124045636106584174</guid><pubDate>Sat, 21 Nov 2026 07:41:57 GMT</pubDate><description>这是我今天读到的最美的一段文字

https://x.com/livelikebruce/status/2123821553444155458?a=IhcghvGRcBxgqiUL_rU&amp;s=46</description></item><item><title>@op7418: 最近从  Fable 5.2 开始，到 GPT-6.2 Astra，再到最新的 Gemini Neon

感觉 AI 做游戏已经是 Vibe Coding 之后另一个非常稳定且很有前景的用例了。

刚</title><link>https://x.com/op7418/status/2124041210805600332</link><guid isPermaLink="false">2124041210805600332</guid><pubDate>Sat, 21 Nov 2026 07:24:22 GMT</pubDate><description>最近从  Fable 5.2 开始，到 GPT-6.2 Astra，再到最新的 Gemini Neon

感觉 AI 做游戏已经是 Vibe Coding 之后另一个非常稳定且很有前景的用例了。

刚好我朋友一直在做 AI 游戏工具 Yoroll，而且他们平台现在还有 4 折的 Fable 5.2 可以用。

他们这个产品内置了非常多 AI 做游戏需要的工具：

像 3D 素材模型、角色动作片段、音乐和音效都可以直接调用，不用每次都让模型重新生成，大幅减少了 token 消耗。

游戏需要在不同客户端和平台做适配，他们平台内置了 HTML5、Godot、Unity 等多种不同引擎。你不需要自己去安装、运行和配置引擎，在软件内就能直接调试。

针对网页游戏做了大量优化，内置很多工具帮你让游戏在网页里顺畅跑起来。

最顶的是在发布上做了很多工作，支持一键发布到 Google Play。以往对很多开发者来说挺难搞的变现环节，现在做出来的小游戏能直接拿来赚钱。</description></item><item><title>@oran_ge: 相比 AI 证明黎曼猜想
1859年提出黎曼猜想的那个人类更令人惊叹啊</title><link>https://x.com/oran_ge/status/2124036780135780840</link><guid isPermaLink="false">2124036780135780840</guid><pubDate>Sat, 21 Nov 2026 07:06:45 GMT</pubDate><description>相比 AI 证明黎曼猜想
1859年提出黎曼猜想的那个人类更令人惊叹啊</description></item><item><title>@hitw93: 良渚的一麵面馆好吃的
来这边玩小伙伴一定要来试试
虾爆鳝好香的！</title><link>https://x.com/hitw93/status/2124024198356189484</link><guid isPermaLink="false">2124024198356189484</guid><pubDate>Sat, 21 Nov 2026 06:16:46 GMT</pubDate><description>良渚的一麵面馆好吃的
来这边玩小伙伴一定要来试试
虾爆鳝好香的！</description></item><item><title>@thsottiaux: We’ve been messing around a bit and inadvertently created an experience that’s a bit too fun interna</title><link>https://x.com/thsottiaux/status/2123997735326232775</link><guid isPermaLink="false">2123997735326232775</guid><pubDate>Sat, 21 Nov 2026 04:31:36 GMT</pubDate><description>We’ve been messing around a bit and inadvertently created an experience that’s a bit too fun internally and now everyone is just having fun instead of working. We might not recover from this.</description></item><item><title>@hitw93: 又整了一个充电头，大疆 140w，还多买了一根线，大疆的一些产品让我挺惊喜的的，我之前买过一个安克的 140w，有时候电流声大的有点吓人还不让退货，然后酷科态那款查了查安全风评一般。

最后就考虑这个</title><link>https://x.com/hitw93/status/2123975928313418072</link><guid isPermaLink="false">2123975928313418072</guid><pubDate>Sat, 21 Nov 2026 03:04:57 GMT</pubDate><description>又整了一个充电头，大疆 140w，还多买了一根线，大疆的一些产品让我挺惊喜的的，我之前买过一个安克的 140w，有时候电流声大的有点吓人还不让退货，然后酷科态那款查了查安全风评一般。

最后就考虑这个了，这次试试他家充电头看看怎么样，有没有买过小伙伴，欢迎推荐好用的充电头！</description></item><item><title>@trq212: I&apos;m calling it Home Ground.</title><link>https://x.com/trq212/status/2123909358828539965</link><guid isPermaLink="false">2123909358828539965</guid><pubDate>Fri, 20 Nov 2026 22:40:26 GMT</pubDate><description>I&apos;m calling it Home Ground.</description></item><item><title>@trq212: I think I’ve figured out the gameplay loop!

It’s a sports anime game, but the sport is elemental be</title><link>https://x.com/trq212/status/2123909357071159327</link><guid isPermaLink="false">2123909357071159327</guid><pubDate>Fri, 20 Nov 2026 22:40:25 GMT</pubDate><description>I think I’ve figured out the gameplay loop!

It’s a sports anime game, but the sport is elemental bending.

You fight in tournaments alongside your squad, and between matches they have their own stories you can get into, like in Mass Effect or Dragon Age.</description></item><item><title>@oran_ge: 很难想象网易暴跌的归因是红果短剧吧
但仔细想想也有道理

字节是时间之神 
游戏也在劫难逃</title><link>https://x.com/oran_ge/status/2123897506706911336</link><guid isPermaLink="false">2123897506706911336</guid><pubDate>Fri, 20 Nov 2026 21:53:20 GMT</pubDate><description>很难想象网易暴跌的归因是红果短剧吧
但仔细想想也有道理

字节是时间之神 
游戏也在劫难逃</description></item><item><title>@alexandr_wang: 🚨 MUSE FEATURE ALERT!!! 🚨

muse can turn the pics in your camera into product shots that actually </title><link>https://x.com/alexandr_wang/status/2123883550382674150</link><guid isPermaLink="false">2123883550382674150</guid><pubDate>Fri, 20 Nov 2026 20:57:52 GMT</pubDate><description>🚨 MUSE FEATURE ALERT!!! 🚨

muse can turn the pics in your camera into product shots that actually sell 📸 ✨ 

take a shot of whatever you’re trying to get rid of and muse makes it look expensive</description></item><item><title>@sundarpichai: Somewhere Euler, Gauss and Riemann in awe today!</title><link>https://x.com/sundarpichai/status/2123869629164786137</link><guid isPermaLink="false">2123869629164786137</guid><pubDate>Fri, 20 Nov 2026 20:02:33 GMT</pubDate><description>Somewhere Euler, Gauss and Riemann in awe today!</description></item><item><title>@thsottiaux: First tweet from Akshay. Be nice to this one, he’s special.</title><link>https://x.com/thsottiaux/status/2123864112694509995</link><guid isPermaLink="false">2123864112694509995</guid><pubDate>Fri, 20 Nov 2026 19:40:38 GMT</pubDate><description>First tweet from Akshay. Be nice to this one, he’s special.</description></item><item><title>@thsottiaux: Team is locked in, upgrade your desktop app. Faster, simpler and a lot more pleasant to use.</title><link>https://x.com/thsottiaux/status/2123861214644347122</link><guid isPermaLink="false">2123861214644347122</guid><pubDate>Fri, 20 Nov 2026 19:29:07 GMT</pubDate><description>Team is locked in, upgrade your desktop app. Faster, simpler and a lot more pleasant to use.</description></item><item><title>@alexandr_wang: legal work is a demanding test for AI

excited to see muse wonder perform so well on @wearelegora’s </title><link>https://x.com/alexandr_wang/status/2123837988920856737</link><guid isPermaLink="false">2123837988920856737</guid><pubDate>Fri, 20 Nov 2026 17:56:50 GMT</pubDate><description>legal work is a demanding test for AI

excited to see muse wonder perform so well on @wearelegora’s BAR benchmark. huge credit to the team!</description></item><item><title>@vista8: 深圳GEO大会，朋友们明天见。

志愿者朋友们辛苦了！</title><link>https://x.com/vista8/status/2123812303024140457</link><guid isPermaLink="false">2123812303024140457</guid><pubDate>Fri, 20 Nov 2026 16:14:46 GMT</pubDate><description>深圳GEO大会，朋友们明天见。

志愿者朋友们辛苦了！</description></item><item><title>@oran_ge: gpt 的审美太差了
包括对文字的审美
用了carbon真的能感觉到文字之美</title><link>https://x.com/oran_ge/status/2123792828988019130</link><guid isPermaLink="false">2123792828988019130</guid><pubDate>Fri, 20 Nov 2026 14:57:23 GMT</pubDate><description>gpt 的审美太差了
包括对文字的审美
用了carbon真的能感觉到文字之美</description></item><item><title>@lifesinger: https://x.com/i/article/2123643516983369728</title><link>https://x.com/lifesinger/status/2123643556044607930</link><guid isPermaLink="false">2123643556044607930</guid><pubDate>Fri, 20 Nov 2026 05:04:13 GMT</pubDate><description>https://x.com/i/article/2123643516983369728</description></item><item><title>@khazix0918: https://x.com/i/article/2123626516731105280</title><link>https://x.com/khazix0918/status/2123627885730951241</link><guid isPermaLink="false">2123627885730951241</guid><pubDate>Fri, 20 Nov 2026 04:01:57 GMT</pubDate><description>https://x.com/i/article/2123626516731105280</description></item><item><title>@vista8: 明天我和姚老师  @yaojingang 组织的深圳 GEO 就要召开了。

除了第三个组，全部参加了本次 GEO 大会，哈哈哈。

活动组织靠热心志愿者，非常给力。

会议酒店旁就是美食一条街，白天</title><link>https://x.com/vista8/status/2123598467318874178</link><guid isPermaLink="false">2123598467318874178</guid><pubDate>Fri, 20 Nov 2026 02:05:03 GMT</pubDate><description>明天我和姚老师  @yaojingang 组织的深圳 GEO 就要召开了。

除了第三个组，全部参加了本次 GEO 大会，哈哈哈。

活动组织靠热心志愿者，非常给力。

会议酒店旁就是美食一条街，白天开会学习，晚上喝酒聊天，想想就很美！</description></item><item><title>@alexandr_wang: 🚨 MUSE FEATURE EXPANSION ALERT!!! 🚨

email is now rolling out to everyone in supported markets 💌 </title><link>https://x.com/alexandr_wang/status/2123596935231856931</link><guid isPermaLink="false">2123596935231856931</guid><pubDate>Fri, 20 Nov 2026 01:58:58 GMT</pubDate><description>🚨 MUSE FEATURE EXPANSION ALERT!!! 🚨

email is now rolling out to everyone in supported markets 💌 

pick your @muse.ai email handle and get started!</description></item><item><title>@lifesinger: 推荐下知识星球创始人吴鲁加的新书《夹缝生长：小产品创业实践》。

刚读了一半，感觉特别适合 AI 时代的超级个体和中小创业者。里面全是大实话，比大实话更吸引人的，是非常多的案例分析，诱人深思，促人行动</title><link>https://x.com/lifesinger/status/2123592344561967492</link><guid isPermaLink="false">2123592344561967492</guid><pubDate>Fri, 20 Nov 2026 01:40:44 GMT</pubDate><description>推荐下知识星球创始人吴鲁加的新书《夹缝生长：小产品创业实践》。

刚读了一半，感觉特别适合 AI 时代的超级个体和中小创业者。里面全是大实话，比大实话更吸引人的，是非常多的案例分析，诱人深思，促人行动。

比如谈及初创公司的优势时，有一个大道理是：做不可扩展的事。其他书很可能停留在 Paul Graham 的这句话上。这本书却给出了具体的操作建议：

1、创始人手动一个个去拉最初的付费用户
2、给用户一张张手写感谢卡
3、一对一教用户操作产品

同时指出：如果某件事情容易规模化，那大厂更适合做。比如做广告、铺渠道、做品牌等。大厂更有钱、有人、有资源、有品牌，在可规模化的事上，创业者很难打败大公司。

不可规模化的事，大厂去做很不划算，就往往不会去做。创业公司去做，反而有新机会。</description></item><item><title>@alexandr_wang: download link here: https://chromewebstore.google.com/detail/muse-from-meta/jnakhhjmpmoohbjgfccbccfo</title><link>https://x.com/alexandr_wang/status/2123591820512993532</link><guid isPermaLink="false">2123591820512993532</guid><pubDate>Fri, 20 Nov 2026 01:38:39 GMT</pubDate><description>download link here: https://chromewebstore.google.com/detail/muse-from-meta/jnakhhjmpmoohbjgfccbccfoifnmbdjd?pli=1</description></item><item><title>@oran_ge: 今日金句：
AI 是最强大的学习工具
AI 也是最强大的逃避学习的工具</title><link>https://x.com/oran_ge/status/2123582386395091186</link><guid isPermaLink="false">2123582386395091186</guid><pubDate>Fri, 20 Nov 2026 01:01:09 GMT</pubDate><description>今日金句：
AI 是最强大的学习工具
AI 也是最强大的逃避学习的工具</description></item><item><title>@lifesinger: 写得真好

成熟的开端就是停止展示自己
反而专注于打磨自己
所有的人都是从我行我素里开始散发出魅力</title><link>https://x.com/lifesinger/status/2123581906726117853</link><guid isPermaLink="false">2123581906726117853</guid><pubDate>Fri, 20 Nov 2026 00:59:15 GMT</pubDate><description>写得真好

成熟的开端就是停止展示自己
反而专注于打磨自己
所有的人都是从我行我素里开始散发出魅力</description></item><item><title>@vista8: 一键高光剪辑：老罗访谈陈晓卿，如何找优秀餐馆片段。

方法也挺符合常识。</title><link>https://x.com/vista8/status/2123560679643009272</link><guid isPermaLink="false">2123560679643009272</guid><pubDate>Thu, 19 Nov 2026 23:34:54 GMT</pubDate><description>一键高光剪辑：老罗访谈陈晓卿，如何找优秀餐馆片段。

方法也挺符合常识。</description></item><item><title>@sundarpichai: Read more on the Google Research Blog: https://research.google/blog/a-game-you-cannot-win-unless-you</title><link>https://x.com/sundarpichai/status/2123529357767672000</link><guid isPermaLink="false">2123529357767672000</guid><pubDate>Thu, 19 Nov 2026 21:30:26 GMT</pubDate><description>Read more on the Google Research Blog: https://research.google/blog/a-game-you-cannot-win-unless-you-have-a-quantum-computer/</description></item><item><title>@sundarpichai: Entanglement usually gets top billing in quantum computing, but quantum contextuality, the idea that</title><link>https://x.com/sundarpichai/status/2123529356089946543</link><guid isPermaLink="false">2123529356089946543</guid><pubDate>Thu, 19 Nov 2026 21:30:26 GMT</pubDate><description>Entanglement usually gets top billing in quantum computing, but quantum contextuality, the idea that properties don&apos;t have fixed values independent of how you measure them, may be the deeper reason quantum computers can outperform classical ones.

Today on the cover of @ScienceMagazine, we put contextuality to work on our Willow quantum processor:

- won the &quot;impossible&quot; magic square game 98.3% of the time (classical limit: 88.9%)
- measured a 5.59 on state-independent contextuality tests (classical bound: 4, quantum max: 6)
- beat classical strategies in multiplayer games up to 71 qubits
- ran a landmark shallow-circuit algorithm at scale for the first time - up to 105 qubits

These tasks also catch subtle hardware errors random benchmarks miss, a sharper ruler for building better quantum systems. https://x.com/ScienceMagazine/status/2123492766214668527/photo/1</description></item><item><title>@vista8: 用 Gemini Neon 给 Subcast 开发了一个视频高光剪辑功能。

发现剪辑速度和封面设计稍差，交给 Opus5.5 优化中。

论代码开发效率，目前还是 Opus 5.5 胜出。

希望</title><link>https://x.com/vista8/status/2123483659592405036</link><guid isPermaLink="false">2123483659592405036</guid><pubDate>Thu, 19 Nov 2026 18:28:51 GMT</pubDate><description>用 Gemini Neon 给 Subcast 开发了一个视频高光剪辑功能。

发现剪辑速度和封面设计稍差，交给 Opus5.5 优化中。

论代码开发效率，目前还是 Opus 5.5 胜出。

希望明天忙里偷闲能搞定这个功能并发布。</description></item><item><title>@lifesinger: 居然提前两天准备好了 GEO 大会的分享 PPT。内容都是我写的，风格偷了个懒，交给了 YouMind 直出。

写文章这事，还是能回到手搓。并且手搓的快乐，确实很难用 AI 替代。

AI 让内容创</title><link>https://x.com/lifesinger/status/2123400216665849890</link><guid isPermaLink="false">2123400216665849890</guid><pubDate>Thu, 19 Nov 2026 12:57:17 GMT</pubDate><description>居然提前两天准备好了 GEO 大会的分享 PPT。内容都是我写的，风格偷了个懒，交给了 YouMind 直出。

写文章这事，还是能回到手搓。并且手搓的快乐，确实很难用 AI 替代。

AI 让内容创作门槛快速降低，同时 AI 直出的质量往往也不错，这使得：

1、创意无比重要。有点像红果短剧，真正制约红果发展的，是番茄小说的剧情。

2、现场无比重要。现场包括时间感和空间感，都是 AI 缺失的。AI 不会一觉醒来，恍如梦中。AI 也体感不到 GEO 大会里，参会创业者们眼角的兴奋或心底的焦虑。

AI 很好。比 AI 更好的，是做人。

下图是周日要在 GEO 大会分享的其中一页，欢迎来现场，或预约线上直播。

欢迎用 YouMind 做 PPT，让自己可以聚焦分享内容本身，其他交给 AI 搞定。</description></item><item><title>@khazix0918: Google：沟通极舒服的模型+审美不错+粗心大意+狗屎的Harness+随便用
Claude：沟通还不错的模型+审美极佳+一般严谨+无敌的Harness+就是不给你用你打我啊略略略
GPT：没法沟通</title><link>https://x.com/khazix0918/status/2123286419158536474</link><guid isPermaLink="false">2123286419158536474</guid><pubDate>Thu, 19 Nov 2026 05:25:05 GMT</pubDate><description>Google：沟通极舒服的模型+审美不错+粗心大意+狗屎的Harness+随便用
Claude：沟通还不错的模型+审美极佳+一般严谨+无敌的Harness+就是不给你用你打我啊略略略
GPT：没法沟通的模型+审美勾槽的+干活极其严谨+还不错的Harness+随便用

所以，啥时候我们才能有一个真正的版本答案呢。。。</description></item><item><title>@khazix0918: 我已经被Google弄得选择困难症发作了...
这是我觉得最混乱的命名...</title><link>https://x.com/khazix0918/status/2123271194841584096</link><guid isPermaLink="false">2123271194841584096</guid><pubDate>Thu, 19 Nov 2026 04:24:35 GMT</pubDate><description>我已经被Google弄得选择困难症发作了...
这是我觉得最混乱的命名...</description></item><item><title>@sundarpichai: Really neat part of Genie 4 is you can type new prompts while exploring a generated world, like drop</title><link>https://x.com/sundarpichai/status/2123241253982916795</link><guid isPermaLink="false">2123241253982916795</guid><pubDate>Thu, 19 Nov 2026 02:25:37 GMT</pubDate><description>Really neat part of Genie 4 is you can type new prompts while exploring a generated world, like dropping in characters or changing the scene + the native audio adapts in real time! Now available in Project Genie for eligible Google AI Ultra subscribers.</description></item><item><title>@sundarpichai: Below Neon 3D modeled a photorealistic lab simulation in Antigravity via Blender MCP 🔥</title><link>https://x.com/sundarpichai/status/2123184309137965524</link><guid isPermaLink="false">2123184309137965524</guid><pubDate>Wed, 18 Nov 2026 22:39:20 GMT</pubDate><description>Below Neon 3D modeled a photorealistic lab simulation in Antigravity via Blender MCP 🔥</description></item><item><title>@satyanadella: Bringing Salesforce to Copilot as we build Copilot as the OS for work!</title><link>https://x.com/satyanadella/status/2123119560803176533</link><guid isPermaLink="false">2123119560803176533</guid><pubDate>Wed, 18 Nov 2026 18:22:03 GMT</pubDate><description>Bringing Salesforce to Copilot as we build Copilot as the OS for work!</description></item><item><title>@sama: So fun to do DealBook with my friend and partner of 11 years, @gdb. Thrilled the world got to see th</title><link>https://x.com/sama/status/2123103873166381287</link><guid isPermaLink="false">2123103873166381287</guid><pubDate>Wed, 18 Nov 2026 17:19:43 GMT</pubDate><description>So fun to do DealBook with my friend and partner of 11 years, @gdb. Thrilled the world got to see the brilliant, fiercely determined leader I have the privilege of working with every day.

Greg’s role at OpenAI has evolved from writing code 100 hours a week in the early days to now leading some of the hardest and most consequential parts of our business.

One thing that has not changed is his extraordinary commitment to our mission. I can say with confidence that OpenAI would not exist without Greg, and I believe the world would be on a much worse path toward the AI future without him.</description></item><item><title>@lifesinger: AI 应用创业者的几个阶段：

第一阶段：卷产品
第二阶段：卷增长
第三阶段：卷产品
第四阶段：卷增长
…
第九阶段：卷组织
第十阶段：卷资本

也有直接从第十阶段开始往第一阶段发展的。你在第几阶段。</title><link>https://x.com/lifesinger/status/2122919071771681120</link><guid isPermaLink="false">2122919071771681120</guid><pubDate>Wed, 18 Nov 2026 05:05:23 GMT</pubDate><description>AI 应用创业者的几个阶段：

第一阶段：卷产品
第二阶段：卷增长
第三阶段：卷产品
第四阶段：卷增长
…
第九阶段：卷组织
第十阶段：卷资本

也有直接从第十阶段开始往第一阶段发展的。你在第几阶段。</description></item><item><title>@satyanadella: Highlights from my opening of the Ignite keynote this morning on how we’re building a frontier ecosy</title><link>https://x.com/satyanadella/status/2122874904903397495</link><guid isPermaLink="false">2122874904903397495</guid><pubDate>Wed, 18 Nov 2026 02:09:53 GMT</pubDate><description>Highlights from my opening of the Ignite keynote this morning on how we’re building a frontier ecosystem to turn advances in AI into real-world impact.</description></item><item><title>@khazix0918: https://x.com/i/article/2122855388030877697</title><link>https://x.com/khazix0918/status/2122858329248825669</link><guid isPermaLink="false">2122858329248825669</guid><pubDate>Wed, 18 Nov 2026 01:04:01 GMT</pubDate><description>https://x.com/i/article/2122855388030877697</description></item><item><title>@satyanadella: With Meta’s latest models coming to Foundry, we’re giving developers and enterprises even more choic</title><link>https://x.com/satyanadella/status/2122800683867082960</link><guid isPermaLink="false">2122800683867082960</guid><pubDate>Tue, 17 Nov 2026 21:14:57 GMT</pubDate><description>With Meta’s latest models coming to Foundry, we’re giving developers and enterprises even more choice and flexibility!</description></item><item><title>@sama: Although the new models are very smart, we have more work ahead of us on making the models nice to w</title><link>https://x.com/sama/status/2122798042734633452</link><guid isPermaLink="false">2122798042734633452</guid><pubDate>Tue, 17 Nov 2026 21:04:27 GMT</pubDate><description>Although the new models are very smart, we have more work ahead of us on making the models nice to work with, and will prioritize that for upcoming releases.

But even if it takes a little extra work, the intelligence and capability are extraordinary.</description></item><item><title>@sama: 6.2 Sol and Astra are out.

These models can do amazing things and are very token-efficient; we hope</title><link>https://x.com/sama/status/2122798039899263456</link><guid isPermaLink="false">2122798039899263456</guid><pubDate>Tue, 17 Nov 2026 21:04:27 GMT</pubDate><description>6.2 Sol and Astra are out.

These models can do amazing things and are very token-efficient; we hope people will get great value from them.

In particular, 6.2 Sol ultrafast is something I think people will really enjoy.</description></item><item><title>@satyanadella: This is the promise of the frontier ecosystem. Read more: https://news.microsoft.com/events/ignite/2</title><link>https://x.com/satyanadella/status/2122793055740985797</link><guid isPermaLink="false">2122793055740985797</guid><pubDate>Tue, 17 Nov 2026 20:44:38 GMT</pubDate><description>This is the promise of the frontier ecosystem. Read more: https://news.microsoft.com/events/ignite/2026/live-news/</description></item><item><title>@satyanadella: Plus, we announced new partnerships and shared work we are doing with customers in every industry se</title><link>https://x.com/satyanadella/status/2122793054147084581</link><guid isPermaLink="false">2122793054147084581</guid><pubDate>Tue, 17 Nov 2026 20:44:38 GMT</pubDate><description>Plus, we announced new partnerships and shared work we are doing with customers in every industry sector, including @AlaskaAir, @ATT, @BlackRock, @BNYglobal, @Chevron, @HeyEpic, @PublicisGroupe, @salesforce, @Unilever, and more. 

Ultimately, the key imperatives for us are to ensure human ambition, a firm&apos;s control of its own intelligence, and trust in technology that gets broadly diffused is maintained.</description></item><item><title>@drfeifei: AI is a civilizational technology! There are tremendous opportunities and responsibilities for all o</title><link>https://x.com/drfeifei/status/2122377911873491397</link><guid isPermaLink="false">2122377911873491397</guid><pubDate>Mon, 16 Nov 2026 17:15:00 GMT</pubDate><description>AI is a civilizational technology! There are tremendous opportunities and responsibilities for all of us at this moment in history. I&apos;m deeply grateful and excited to lead the @theworldlabs team to join @AMD to embark on a new beginning, to advance science, develop technologies, and ultimately make a positive impact to the world! 😍🚀</description></item><item><title>@drfeifei: Yes! Come and join us to build better technologies and more importantly, a better world together! We</title><link>https://x.com/drfeifei/status/2122376424149721342</link><guid isPermaLink="false">2122376424149721342</guid><pubDate>Mon, 16 Nov 2026 17:09:06 GMT</pubDate><description>Yes! Come and join us to build better technologies and more importantly, a better world together! We promise a thrilling journey to work on crazy hard problems, and to work with ambitious, curious, intellectually honest, courageous and very kind-hearted and fun teams! 🚀</description></item><item><title>@drfeifei: AI is a civilizational technology! I&apos;m deeply grateful and excited by the opportunities and responsi</title><link>https://x.com/drfeifei/status/2122375532163866756</link><guid isPermaLink="false">2122375532163866756</guid><pubDate>Mon, 16 Nov 2026 17:05:33 GMT</pubDate><description>AI is a civilizational technology! I&apos;m deeply grateful and excited by the opportunities and responsibility awaiting all of us! Leading @theworldlabs for the past 2+ years has been an honor of my life. I cannot wait to continue working with this incredible team of innovators and scientists, and also to embark on this new chapter with all the colleagues at @AMD ! Lets make the world a better place for everyone! 😍🚀</description></item><item><title>@drfeifei: Thank you @LisaSu! There is so much ahead of us to create the best technologies to benefit everyone!</title><link>https://x.com/drfeifei/status/2122340036645621862</link><guid isPermaLink="false">2122340036645621862</guid><pubDate>Mon, 16 Nov 2026 14:44:30 GMT</pubDate><description>Thank you @LisaSu! There is so much ahead of us to create the best technologies to benefit everyone! The entire @theworldlabs team and I are beyond excited! 🤩🚀</description></item><item><title>@tianyi: 下一个超级 App 不必是面向所有人的一个超级 App。

现在，每个人都有能力创造一个超级 App。

下一个超级 App 也可以是只属于你一个人的超级 App。</title><link>https://x.com/tianyi/status/2121264664571269160</link><guid isPermaLink="false">2121264664571269160</guid><pubDate>Fri, 13 Nov 2026 15:31:21 GMT</pubDate><description>下一个超级 App 不必是面向所有人的一个超级 App。

现在，每个人都有能力创造一个超级 App。

下一个超级 App 也可以是只属于你一个人的超级 App。</description></item><item><title>@tianyi: DeepSeek 产品背后的各项核心能力均已通过 API 开放。我们期待在产品功能的实现门槛进一步降低之后，能够从社区中生长出更多样、更极致、更具想象力的产品！</title><link>https://x.com/tianyi/status/2121248749343420437</link><guid isPermaLink="false">2121248749343420437</guid><pubDate>Fri, 13 Nov 2026 14:28:07 GMT</pubDate><description>DeepSeek 产品背后的各项核心能力均已通过 API 开放。我们期待在产品功能的实现门槛进一步降低之后，能够从社区中生长出更多样、更极致、更具想象力的产品！</description></item><item><title>@yetone: 唉，大家可能不知道，它是有点名表扬机制的。</title><link>https://x.com/yetone/status/2120439802432749914</link><guid isPermaLink="false">2120439802432749914</guid><pubDate>Wed, 11 Nov 2026 08:53:39 GMT</pubDate><description>唉，大家可能不知道，它是有点名表扬机制的。</description></item><item><title>@yetone: 昨天这么多的吗</title><link>https://x.com/yetone/status/2120400776170500431</link><guid isPermaLink="false">2120400776170500431</guid><pubDate>Wed, 11 Nov 2026 06:18:34 GMT</pubDate><description>昨天这么多的吗</description></item><item><title>@yetone: 看到 Alma 了，激动帮转</title><link>https://x.com/yetone/status/2120196856416452631</link><guid isPermaLink="false">2120196856416452631</guid><pubDate>Tue, 10 Nov 2026 16:48:16 GMT</pubDate><description>看到 Alma 了，激动帮转</description></item><item><title>@yetone: 看来是把一个像我一样喜欢复古游戏机的人招过去了，破坏了大局</title><link>https://x.com/yetone/status/2120061343290282167</link><guid isPermaLink="false">2120061343290282167</guid><pubDate>Tue, 10 Nov 2026 07:49:47 GMT</pubDate><description>看来是把一个像我一样喜欢复古游戏机的人招过去了，破坏了大局</description></item><item><title>@yetone: 你们老是叫我叶天帝，甚至有人还叫我老叶，导致我现在刷推，老是觉得我盗了一个姓叶的人的账号在刷推，这种配得感完全丧失的感觉真的很奇妙</title><link>https://x.com/yetone/status/2120061179112636506</link><guid isPermaLink="false">2120061179112636506</guid><pubDate>Tue, 10 Nov 2026 07:49:08 GMT</pubDate><description>你们老是叫我叶天帝，甚至有人还叫我老叶，导致我现在刷推，老是觉得我盗了一个姓叶的人的账号在刷推，这种配得感完全丧失的感觉真的很奇妙</description></item><item><title>@bcherny: Pro users get one reset, and Max, Ultra, and Team Premium get up to three, depending on how long you</title><link>https://x.com/bcherny/status/2119890820186976763</link><guid isPermaLink="false">2119890820186976763</guid><pubDate>Mon, 09 Nov 2026 20:32:11 GMT</pubDate><description>Pro users get one reset, and Max, Ultra, and Team Premium get up to three, depending on how long you&apos;ve subscribed:
https://support.claude.com/en/articles/17506494-claude-fable-5-2-launch-offers</description></item><item><title>@bcherny: We wanted to do more for the people who’ve been with us the longest. So with Fable 5.2, the longer y</title><link>https://x.com/bcherny/status/2119873095129550880</link><guid isPermaLink="false">2119873095129550880</guid><pubDate>Mon, 09 Nov 2026 19:21:45 GMT</pubDate><description>We wanted to do more for the people who’ve been with us the longest. So with Fable 5.2, the longer you’ve been subscribed, the more Fable resets you get.</description></item><item><title>@tianyi: 很多 DeepSeek Harness 插件作者都有问过我，什么样的插件和功能官方会做成内置，如果一个功能官方会做，那第三方插件做还是否有意义。

我们的希望是，DeepSeek Harness 可以</title><link>https://x.com/tianyi/status/2119433744671273441</link><guid isPermaLink="false">2119433744671273441</guid><pubDate>Sun, 08 Nov 2026 14:15:56 GMT</pubDate><description>很多 DeepSeek Harness 插件作者都有问过我，什么样的插件和功能官方会做成内置，如果一个功能官方会做，那第三方插件做还是否有意义。

我们的希望是，DeepSeek Harness 可以作为 Agent 时代的操作系统。

所以，也许可以用操作系统来举例。以智能手机操作系统为例，系统会内置一些通用的 App，例如闹铃与计算器，满足用户的基本需求。但做手机操作系统的厂商肯定更希望生态中能出现更多更广的 App 满足更广泛的需求，希望智能手机生态中出现更好的更适合用户的闹铃 App 与计算器 App 能够比内置功能做得更好，希望第三方开发者能够使用智能手机操作系统提供的 API 做出发明智能手机的人从来没想过的 App。

DeepSeek Harness 对于第三方插件生态的想法会是类似的。我们会为用户提供稳定的基础能力，并内置大多数人都需要的通用功能。但在“一切皆插件”的设计下，内置功能同样可以被第三方插件替换，我们希望更多的社区作品成为用户的日常选择。</description></item><item><title>@karpathy: Earlier this week Tristan published his top YouTube channels:
https://thume.ca/2026/11/01/youtube-ti</title><link>https://x.com/karpathy/status/2119114537442189652</link><guid isPermaLink="false">2119114537442189652</guid><pubDate>Sat, 07 Nov 2026 17:07:31 GMT</pubDate><description>Earlier this week Tristan published his top YouTube channels:
https://thume.ca/2026/11/01/youtube-tier-list-2026/
Which looked excellent so clearly I had to turn it into TristanTube:
https://karpathy.ai/tristan-tube/
47K super nerdy videos +TV mode for laid back watching. (One-shotted with Claude Code in 3 minutes)</description></item><item><title>@AndrewYNg: Join me and 2000 other AI engineers and builders at AI Dev in New York City, Nov 30-Dec 1!

A Linked</title><link>https://x.com/AndrewYNg/status/2118069849142558737</link><guid isPermaLink="false">2118069849142558737</guid><pubDate>Wed, 04 Nov 2026 19:56:18 GMT</pubDate><description>Join me and 2000 other AI engineers and builders at AI Dev in New York City, Nov 30-Dec 1!

A LinkedIn study found that over July-September, the job title AI Engineer has been one of the 10 fastest-growing occupations in seven major economies. AI engineering has become a core skill for all software work, and AI Dev is designed to help you develop it.

Our sessions give you hands-on practice with:
- Coding agents and harness design: Harness tuning, repository context, autonomous coding
- Agent memory and knowledge graphs: GraphRAG, vector search, persistent memory, hybrid retrieval
- Agents in production: Evals, observability, simulated testing, browser and voice agents

@ylecun and I will also talk about what’s next in world models and physical AI.

I hope to see you there!

Tickets: https://ai-dev.deeplearning.ai/</description></item><item><title>@jietang: GLM-5.3 is on T-Systems now....</title><link>https://x.com/jietang/status/2118001280048345450</link><guid isPermaLink="false">2118001280048345450</guid><pubDate>Wed, 04 Nov 2026 15:23:50 GMT</pubDate><description>GLM-5.3 is on T-Systems now....</description></item><item><title>@sama: People should always be in control of the future.

OpenAI is against the notion that it would be acc</title><link>https://x.com/sama/status/2117794005165748448</link><guid isPermaLink="false">2117794005165748448</guid><pubDate>Wed, 04 Nov 2026 01:40:12 GMT</pubDate><description>People should always be in control of the future.

OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”.

People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.</description></item><item><title>@bcherny: Signing in to your company&apos;s plugins one by one is annoying, so we let Claude auto-connect to them n</title><link>https://x.com/bcherny/status/2117693181651542387</link><guid isPermaLink="false">2117693181651542387</guid><pubDate>Tue, 03 Nov 2026 18:59:33 GMT</pubDate><description>Signing in to your company&apos;s plugins one by one is annoying, so we let Claude auto-connect to them now.

Team and Enterprise admins can set them up, and teammates are connected automatically the first time they use the plugins.</description></item><item><title>@tianyi: Have you tried the Creator Mode in @DeepSeekHarness? Any feedback?</title><link>https://x.com/tianyi/status/2117258349649780737</link><guid isPermaLink="false">2117258349649780737</guid><pubDate>Mon, 02 Nov 2026 14:11:41 GMT</pubDate><description>Have you tried the Creator Mode in @DeepSeekHarness? Any feedback?</description></item><item><title>@tianyi: I got a lot of inspiration talking to the brightest minds in the fields @badlogicgames and @mitsuhik</title><link>https://x.com/tianyi/status/2117064134961631414</link><guid isPermaLink="false">2117064134961631414</guid><pubDate>Mon, 02 Nov 2026 01:19:57 GMT</pubDate><description>I got a lot of inspiration talking to the brightest minds in the fields @badlogicgames and @mitsuhiko.

At @deepseek_ai we believe in scaling across agent harnesses and aim to provide good support to @pidotdev among all the major harnesses in the world in each generation of our models.</description></item><item><title>@karpathy: Weekend hack: HackerTube. I was annoyed with YouTube&apos;s recommendations so I had Claude filter down 4</title><link>https://x.com/karpathy/status/2116957135661945283</link><guid isPermaLink="false">2116957135661945283</guid><pubDate>Sun, 01 Nov 2026 18:14:46 GMT</pubDate><description>Weekend hack: HackerTube. I was annoyed with YouTube&apos;s recommendations so I had Claude filter down 43M comments over 20 years of HN to 6,454 videos that keep being recommended. *Significantly* better viewing experience, ~2 hours with Opus 5.5:
https://hackertube.org/

- Excellent recs without any logins (state is kept in local browser session)
- Thumbnail is a representative frame from middle of the video instead of the super duper A/B test hyper-optimized user uploads
- Underlying data from the comment analysis as one zip file, point your Claude at it and ask for your own version
- HackerTube Longs ✊

A bit more on the process:
I use Claude Code (desktop app for Mac), Opus 5.5 on Medium effort as my default. I only talk to Claude and have it do everything, the only &quot;manual work&quot; I helped with in tandem was in the browser (reminiscent of MenuGen): 1) I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata, 2) I booted up a VPS box for serving and 3) I bought the domain name. Once the box was up I gave Claude the IP and told it update, secure, harden, deploy and optimize the code, and it did the whole thing.

Many awesome projects that would have normally taken weeks/months even can now be few-shotted for breakfast. What else are we not thinking of.</description></item><item><title>@bcherny: We&apos;re removing the 5-hour limit for verified Ultra accounts.</title><link>https://x.com/bcherny/status/2116222842413142111</link><guid isPermaLink="false">2116222842413142111</guid><pubDate>Fri, 30 Oct 2026 17:36:57 GMT</pubDate><description>We&apos;re removing the 5-hour limit for verified Ultra accounts.</description></item><item><title>@bcherny: Utah gave Claude Code to 34 developers across 10 state agencies. 80% had a working environment withi</title><link>https://x.com/bcherny/status/2115491259381072269</link><guid isPermaLink="false">2115491259381072269</guid><pubDate>Wed, 28 Oct 2026 17:09:54 GMT</pubDate><description>Utah gave Claude Code to 34 developers across 10 state agencies. 80% had a working environment within 30 minutes, 71% reported saving 20+ hours by the end of the four-week pilot.</description></item><item><title>@sama: 6.1 Terra is here! It is good at a lot of things, but we are especially excited for this new era of </title><link>https://x.com/sama/status/2114818620782248024</link><guid isPermaLink="false">2114818620782248024</guid><pubDate>Mon, 26 Oct 2026 20:37:05 GMT</pubDate><description>6.1 Terra is here! It is good at a lot of things, but we are especially excited for this new era of computer use.</description></item><item><title>@DarioAmodei: Thank you for this, @_NathanCalvin. I appreciate and am a fan of the investigative work you’ve done </title><link>https://x.com/DarioAmodei/status/2112732230124818891</link><guid isPermaLink="false">2112732230124818891</guid><pubDate>Wed, 21 Oct 2026 02:26:30 GMT</pubDate><description>Thank you for this, @_NathanCalvin. I appreciate and am a fan of the investigative work you’ve done on behalf of AI safety and believe your comments are coming from a constructive, honest place, and that it benefits the public to have people whose job is to scrutinize the actions of frontier AI companies. I think something would be wrong if companies of the size of Anthropic didn’t get harshly criticized basically every day; strong antibodies against corporate power are a healthy feature of a free society, and especially so when issues of the gravity of the one we’re facing are at play. I also do agree that some of the examples you cite were mistakes by Anthropic.

That said, I disagree with the thrust of your post. At the core, a good fraction of the things you criticize are consequences of Anthropic’s basic theory of change that we have held for many years. Our strategy has always been what I’ve called the “race to the top” — by showing how it’s possible to build the technology responsibly, we can raise the bar for the entire industry. The forces that drive the race to the top are hugely more effective if the company practicing them is at the frontier; that’s one lesson I’ve learned repeatedly. And looking at that strategy as a whole, I’m proud of our track record and stand by it.

I believe that if we take a list of things that Anthropic did in the past year that were pro-safety and commercially costly to us, they far outweigh the list you’ve made, while at the same time I also disagree with some of the items on your list. Here are some things Anthropic has done, in the same period you’ve cited (the last year), which were very costly to us commercially while having strong safety benefits:

1. We played a key role in SB53 (and other state laws), at a time when doing so was politically perilous to the company and AI regulation was very unpopular. We in fact argued for a stronger version of SB53 than was ultimately passed; there were a lot of stakeholders and we were not the limiting factor in the strength of the bill that was passed.

2. We have very strong biological safeguards on our models, because we’ve seen concerning results in bio uplift trials and even seen and blocked potential attempts to misuse our models for bioweapons (as seen in our threat intelligence report from a few weeks ago). This is a very serious threat that we want to have zero tolerance for. But having very low false negatives means having significant false positives (despite months of effort on our part to improve the classifiers), enough so that biologists joke about how restrictive our classifiers are, and competitors sometimes sell against us by touting their more permissive bio capabilities. This has been very painful to us in an area where we care a lot about supporting an ecosystem of Claude users (biomedical science; and it is particularly frustrating to me as a biologist), but it’s the right thing to do and we intend to keep doing it.

3. Similarly, we first had Mythos-class models in February but didn’t make them generally available until June (when we had stronger cyber safeguards in place). We only offered them to a limited set of partners in order to focus on cyberdefense. It would have been much more economically rational to simply release the model to everyone. We likely sacrificed a significant chunk of a several-month model lead. But because of this, many companies secured their cyber infrastructure before the deluge of widely available cyber-capable models. Of course there are ways it could have gone better or been run better, but this was what we were aiming to do.

4. I advocated forcefully and publicly, in my essay Policy on the AI Exponential in June of this year, for the government to have the power to block model releases. Back then this was still outside or maybe on the edge of the Overton window, though I understand the window has moved since then. We were attacked intensely for this.

5. I don’t want to re-ignite conflict, but the events of February 2026 are a strong example. It is clear that our public stance could have seriously damaged the company’s future.

6. We’ve continued to advocate for export controls on chips and chipmaking equipment to China, because we think it allows democracies to negotiate from a position of strength on AI safety agreements. We’ve done this despite problems it’s created for us with industry partners who hold very different views and on whom we are very dependent.

7. As far as I’m aware we were the first, and still only, frontier company to restrict the sale of AI services to PRC-controlled companies, including those operating through third countries (otherwise a major loophole). We’ve therefore forfeited significant short-term business to prevent fueling AI platforms and applications that would help the CCP’s military and intelligence services. We are also cracking down on distillation; generally, distillation is not a major threat to us commercially, but like chip exports it has concerning national security and safety implications, so we expend significant energy and company resources to address it.

8. As you mentioned, we’ve put out a lot of reports, announcements, long model cards, public advocacy, etc, all trying to inform the public about the risks of AI. These involve a level of detail and effort that would be implausible for a purely profit-maximizing company.

9. Just this week, we announced we were putting 5% (and possibly more over time) of the company into a public benefit fund.

All of these decisions came with very substantial costs to us. To be fair, those that involved holding back or restricting access to frontier capabilities also had a broader cost — for example, the public might otherwise have benefitted from faster progress in biology, so I am not saying that our decisions were without tradeoffs — but they were not driven by profit-seeking.

For each decision, you can come up with ulterior motives that could have driven us — I think it is a general truth that any decision, on any topic, can always be portrayed as a nefarious scheme. But most of those end up being pretty farfetched or don’t fully make sense if you actually examine them closely. For example, it is deeply implausible that it would be a shrewd business move to block a large swath of research biology in order to do the work ourselves and compete with our customers; we expect the benefit to the ecosystem to greatly exceed the work Anthropic does itself, and for the two to complement each other. It is not in Anthropic’s interest to be perceived as unfriendly to biologists. Our Life Sciences Verification Program, which gives vetted biology professionals access to our models, also refutes this idea. 

Of course, a different set of people will see this and make a different criticism: that Anthropic shouldn&apos;t be making decisions about who gets access to what in the first place. I agree, in the sense that ideally the government would be making these decisions, and that’s why we advocate for regulation. But for now we are forced to make them, and enabling bioterrorism isn’t a better option.

I think that even if I conceded every item in your list, I would happily hold up mine next to it; I don’t think it’s a close call. That is of course partly subjective and different things are important to different people. Maybe one way you could object is that several of the items above relate to AI misuse and not loss-of-control; if you think the latter is much more important than the former, that might change your calculus. But this is not Anthropic’s view — we view both as serious issues. You can also disagree with our decisions, as many have, but they are consistent.

But I also disagree with at least some of your list. Several items don’t represent a deviation from Anthropic’s strategy or values at all, once we postulate the “race to the top” strategy that requires being a frontier AI company. Your items 1-4 are all simply examples of what it means to compete at the frontier, and we stand by them, both in the sense that they are transparently ordinary things to do as part of that strategy, and that we stand by the strategy as a whole. 

To be fair, there are important questions worth confronting (and that we are confronting) about what to do when the AI-driven acceleration gets fast enough that it is hard for safety to keep up. We are nearing that point and that’s what’s driving some of our thinking about “pacing.” But 1-4 are actions taken before that point and should be perfectly expected based on Anthropic’s strategy as a company. Of course it is possible to disagree with that strategy, but that’s a more fundamental disagreement that you could have had any time over the last few years.

I agree we could have handled your items 7, 8, and 9 better. All fall into the category of “rhetoric or policy communication that wasn’t carefully written or resulting from fast-changing circumstances not propagating from one team to another fast enough.” This kind of thing happens a lot in large organizations (Anthropic is over 5,000 people), and doesn’t even necessarily represent a failure or mistake of any single person. I am of course the CEO, and I take responsibility. The volume of communication from Anthropic is enormous and if these are the worst examples, I think we’re frankly doing pretty well, though I will continue to work hard to make sure we do even better.

Your 5th item (the change to our RSP) is the most substantive item, and people can debate this, but I think that the old, more rigid RSP created a lot of situations where we were technically required to do something that required a lot of effort but in retrospect wasn’t very important for safety, as well as not covering some of the most important things. Our current RSP’s focus on transparency brought things like risk reports and roadmaps, and its emphasis on independent external review laid the groundwork for even more intensive embedded evaluation, all of which I think have been valuable in driving a race to the top. In the original announcement in September 2023, we described the RSP as an “early iteration that we [would] build on” and said that “rapid iteration…will almost certainly be necessary.” I understand that changing plans is always a move that creates justifiable scrutiny — I just don’t think we ever represented that we wouldn’t do so.

Relating to your 6th item (having an FCF in addition to our RSP), I know our intent is to meet the letter and spirit of our regulatory obligations (via our FCF) while also remaining able to experiment with more novel approaches to risk management that go beyond what the law currently requires (via our RSP). I promise to look into it further, as it’s a complex issue and it is possible we are not weighing the considerations in the right way.

Overall, although Anthropic has made plenty of mistakes as an organization — including some of the items on your list — I feel that we have repeatedly taken serious and substantive actions counter to our bottom line and that there is a significant negativity bias distorting assessments like the one you are making above. We’ll always try to do better but, again, I am extremely proud of our track record and stand by it. I will have more to say soon about how we are going to further live our values in this moment of rapid change.</description></item></channel></rss>