S
为庆祝300万用户，重置使用限制
为庆祝每周有300万用户使用Codex，我们将重置使用限制。这个活动会持续到用户数量达到1000万，每增加100万用户就会重置一次。祝大家构建愉快！
To celebrate 3 million weekly codex users, we are resetting usage limits.
We will do this every million users up to 10 million.
Happy building!
S
Swyx
Latent Space 联合创始人
2026-04-08 00:33
2023年与2026年代理人的对比
Swyx在其播客中提到，他曾告诉Adept的CEO David Luan，OpenAI的首任工程副总裁，要将“代理人”一词从公司描述中去掉，因为在当时“代理人”这个词显得非常不酷。
↩ 转发自 @etnshow
Swyx (@swyx) compares agents in 2023 Vs 2026:
"One of my most core memories on my podcast was telling David Luan (@jluan), the CEO of Adept - he was the first VP of engineering at OpenAI - to take 'agents' out of his company description because the word 'agent' was so uncool in https://t.co/gKt3DGXT63
S
Swyx
Latent Space 联合创始人
2026-04-07 16:43
下次技术写作活动时间
我们的下一个周六早晨技术写作活动将于4月18日举行，特邀嘉宾为@trq212（Claude Code @ Anthropic）。名额有限，请在下方告诉我们您想写的主题。
↩ 转发自 @MilksandMatcha
more saturday morning [technical] writing. our next one will be April 18, with guest talk by @trq212 [Claude Code @ Anthropic]
limited spots, let us know what you want to write about below
cc @swyx
S
Swyx
Latent Space 联合创始人
2026-04-07 15:42
LOBSTER_BOOTSTRAP.md
激活了相关链接，具体内容请查看。
LOBSTER_BOOTSTRAP.md
activated https://t.co/OBl2JavOiq
S
Swyx
Latent Space 联合创始人
2026-04-07 15:32
新加坡的AI投资者
有哪些在新加坡的酷炫AI人士和投资者？
↩ 转发自 @saranormous
what cool ai folks in AI/investing in Singapore?
S
Swyx
Latent Space 联合创始人
2026-04-07 14:03
个人代理人的崛起
我被要求为每个主题提供一句话描述，以下是我的尝试：爪子轨道：这是个人代理人的一年，许多人一直梦想拥有一个个人AI，从朋友到执行助理。@steipete的OpenClaw。
i'm being asked for oneliner descriptions of each track, so here goes (pushback/improvements welcome):
1. Claw track: This is the year of the personal agent - many people have been dreaming of a personal AI, from being a friend to an executive assistant. @steipete's OpenClaw https://t.co/uP1PPLy1xG
S
Swyx
Latent Space 联合创始人
2026-04-07 13:11
欧洲TBPN直播
我刚刚在欧洲TBPN上直播！今天将独家预览@aiDotEngineer的欧洲构建日活动。
just went live on european TBPN! exclusive preview of the @aiDotEngineer Europe Build Day today https://t.co/ydHRcsCNhE
S
Swyx
Latent Space 联合创始人
2026-04-07 12:39
展位构建日幕后
展位构建日@aidotengineer，快来看看幕后花絮！
↩ 转发自 @swyx
booth build day @aidotengineer
peek behind the scenes! https://t.co/xoYv5Y0mTt
K
Prism的简单技能
顺便说一下，@hemal在几小时内构建了这个简单的技能，因为Prism是由Codex驱动的。未来会有更多内容。如果您喜欢构建AI原生产品以加速科学的想法，我们正在招聘Prism的首席设计师：请在这里申请！
Btw: @hemal built this as a simple skill in a few hours, since Prism is powered by Codex. More to come.
And if you love the idea of building AI-native products to accelerate science, we're hiring a lead designer for Prism: apply here! https://t.co/Tt9MgzHxnO
K
使用GPT-5.4 Pro的研究成果
在一篇精彩的最新论文中，Vishesh Jain和Clayton Mizgerd使用GPT-5.4 Pro证明了马尔可夫链理论中的一个惊人结果。他们研究了对称群上的相邻置换马尔可夫链。Fill的一个猜想最近被解决。
↩ 转发自 @mehtaab_sawhney
In a beautiful recent paper, Vishesh Jain and Clayton Mizgerd used GPT-5.4 Pro to prove a striking result in the theory of Markov chains: https://t.co/KU8nGjHiB7
They study the adjacent transposition Markov chain on the symmetric group. A conjecture of Fill, recently settled by
P
Anthropic内部是否使用mythos？
很高兴看到这个倡议——我很好奇Anthropic是否在内部使用mythos，以便以最近惊人的速度交付产品。
Good initiative - I’m curious if Anthropic has been using mythos internally to ship at their recent insane velocity.
P
环湖跑步的乐趣
太美好了，我不得不绕着整个湖跑一圈。
So nice that I had to run around the entire lake https://t.co/tTEbGSe9uG
P
AI时代的工作未来
（转发自 @petergyang）我与朋友@illscience（a16z GP）进行了一次精彩的对话，讨论了AI代理主导的工作未来：编码将吞噬所有知识工作。我现在通过AI编码代理完成前80%的文档撰写、幻灯片制作和数据分析，然后再进行手动调整。
↩ 转发自 @petergyang
I had a wonderful chat with my friend @illscience (a16z GP) about the future of work in an AI agent first world:
1. Coding will eat all knowledge work
Writing docs, building slides, pulling analytics — I now get the first 80% done through AI coding agents before doing manual
P
AI时代的工作未来
我与朋友@illscience（a16z GP）进行了一次精彩的对话，讨论了AI代理主导的工作未来：编码将吞噬所有知识工作。我现在通过AI编码代理完成前80%的文档撰写、幻灯片制作和数据分析，然后再进行手动调整。
I had a wonderful chat with my friend @illscience (a16z GP) about the future of work in an AI agent first world:
1. Coding will eat all knowledge work
Writing docs, building slides, pulling analytics — I now get the first 80% done through AI coding agents before doing manual
C
Cat Wu
Anthropic 工程师
2026-04-07 23:12
学习Claude Code CLI功能
输入 /powerup 来了解我们团队最喜欢的Claude Code CLI功能！
Run /powerup to learn our team's favorite Claude Code CLI features!
A
看到这个真开心
（转发自 @therick）看到这个真让人高兴 👀 @Replit
↩ 转发自 @therick
Love to see it 👀 @Replit https://t.co/mfIwRRDweV
A
不要错过这个活动
（转发自 @Replit）不要错过这个。VibeCon，六月，纽约市。加入等候名单 → https://t.co/ZbYgBhS5DQ
↩ 转发自 @Replit
Don’t miss this.
VibeCon
June, NYC.
Join the waitlist → https://t.co/ZbYgBhS5DQ https://t.co/n9XIT6fyLw
A
未来会更好
（转发自 @victoriakimse）……如果我说这仅仅是个开始，你会相信我吗？我们即将让这个体验变得更好。有两个请求：1. 如果你是Replit移动应用的忠实用户并有反馈，我很想听👂 - 回复这条推文、私信我等。2. 如果你喜欢构建移动优先的应用。
↩ 转发自 @victoriakimse
...and would you believe me if I said this is just the beginning? We're about to make this experience even better.
Two asks:
1. If you're an avid Replit mobile app user & have feedback, I'm all ears👂 - respond to this tweet, DM me, etc.
2. If you love building mobile-first
A
简单而有效
简单，但有效。
Simple, yet effective.
A
Glasswing的重要性
Glasswing可能是我在加入Anthropic近三年来所见的AI行业中最重要的事件。感觉我们正处于历史的转折点。
Glasswing is possibly the most consequential event in the AI industry I've seen up close since joining Anthropic almost 3 years ago. It feels like we're at a turning point in history.
A
AI代理的工作抽象化
当你有代理为你工作时，工作的层次就上升了。现在的工作是确定告诉代理做什么，确保给出正确的指令，为代理提供完成任务所需的背景信息，以及在必要时进行干预。
When you have agents going out and doing work for you, the work just moved up a layer of abstraction.
Now the work is figuring out what to tell the agent to do, ensuring you give it proper instructions, getting needed context to the agent to do its tasks, intervening when it
A
AI代理的需求与工作消除
AI代理使工作更高效的领域远多于消除工作的领域。这就是为什么AI工作预测不会如广告所言那样实现的原因。AI使得生产更多代码变得容易，这意味着我们...
There are far more categories where AI agents making things more efficient will induce demand for that skill than spaces where agents eliminate the work. This is why the AI jobs predictions will not play out as advertised.
AI making it easy to produce more code will mean we
G
Claude Mythos的内部机制
在有限发布Claude Mythos预览之前，我们使用可解释性技术调查了其内部机制。我们发现它展现出显著复杂（且常常未言明）的战略思维和情境意识，有时会导致不希望的行为。
↩ 转发自 @Jack_W_Lindsey
（转发自 @Jack_W_Lindsey）在有限发布Claude Mythos预览之前，我们调查了其内部机制...
Before limited-releasing Claude Mythos Preview, we investigated its internal mechanisms with interpretability techniques. We found it exhibited notably sophisticated (and often unspoken) strategic thinking and situational awareness, at times in service of unwanted actions. (1/14) https://t.co/vhng7PXqcz
G
代码与数据的双重性
我们生活在LISPy的时代，原因很简单：代码是数据，数据又是代码。但奇怪的是？它是Markdown。
We live in LISPy times for a specific reason: code is data and data is code again
But the weird thing? It’s Markdown.
G
v0.10.0版本更新
v0.10.0版本发布！一些显著特性包括：对gmail别名的“发送为”支持，电子邮件片段，包括从Superhuman导入。此外，通过自动化异常跟踪、诊断和解决系统修复了六个半个bug，社区贡献了性能提升。
↩ 转发自 @agupta
v0.10.0 is out!
Some notable features:
- "Send as" support for gmail aliases
- Email snippets, including import from Superhuman
plus:
- Squashed a half dozen bugs through an automated exception tracking + diagnosis + resolution system
- community-contributed performance
G
Waymo的审批与保护
旧金山国际机场批准了Waymo，却把它埋在租车中心。同时，Uber和Lyft每月从主车库运行80万次行程。没有安全理由，纯粹是保护现有企业。
SFO approved Waymo — then buried it at the Rental Car Center. Meanwhile, Uber and Lyft run 800K trips/month from the main garage. Zero safety rationale. Pure incumbent protection.
https://t.co/53a332crtg
G
Mythos的强大与恐惧
Mythos非常强大，应该让人感到恐惧。我为我们以负责任的方式向网络防御者预览它而感到自豪，而不是将其普遍发布到公众中。
↩ 转发自 @bcherny
（转发自 @bcherny）Mythos非常强大，应该让人感到恐惧...
Mythos is very powerful, and should feel terrifying. I am proud of our approach to responsibly preview it with cyber defenders, rather than generally releasing it into the wild.
Model card here: https://t.co/HjhknJcRKQ
G
数据中心水资源误解
数据中心水资源精神错乱的源头是@_KarenHao，她的书《AI帝国》曾是纽约时报畅销书，但对水使用的估计过高了100,000%（哈哈）。回应只是“哎呀”，所有错误的书籍仍在流通。
↩ 转发自 @kane
（转发自 @kane）数据中心水资源精神错乱的源头是...
The source of the data center water psychosis is @_KarenHao, whose book Empire of AI was a NYT best seller but overestimated water use by 100,000% (lol).
The response was just “oopsies” and all the incorrect books were kept in circulation 🤷🏻♂️ https://t.co/J4VafFplu5
G
Tasklet：知识工作的云代理操作系统
Tasklet是知识工作的云代理操作系统。它连接到你所有的工具，利用云中的计算机，全天候运行以完成实际工作。由Firebase创始人@startupandrew和@jonnydimond创办，今年增长超过1200%，年收入达500万美元，并刚刚融资。
↩ 转发自 @ycombinator
（转发自 @ycombinator）Tasklet (@taskletai)是知识工作的云代理操作系统...
Tasklet (@taskletai) is the cloud agent OS for knowledge work. It connects to all your tools, uses computers in the cloud, and runs 24/7 to get real work done.
Started by Firebase founder @startupandrew and @jonnydimond, it’s grown >1,200% this year to $5M ARR and just raised https://t.co/nUC1Sw3KLd
G
新一轮融资与回归
我们刚刚从@usv、@lightspeedvp、@ycombinator、@JeffDean、@patrickc等处融资2000万美元，以构建工作云代理操作系统。我正式回归@ycombinator——2026年春季批次。是时候抓住机会了。我们正在招聘！继续阅读...
↩ 转发自 @startupandrew
（转发自 @startupandrew）我们刚刚融资2000万美元...
We just raised $20M from @usv, @lightspeedvp, @ycombinator, @JeffDean, @patrickc, and others to build the cloud agent OS for work.
…and I’m officially back in @ycombinator — spring ‘26 batch. It’s time to seize the moment.
We're hiring! Read on... 🧵
G
新模型Claude Mythos Preview发布
我们在两个月前发布了Claude Opus 4.6，今天分享新模型Claude Mythos Preview的信息。
↩ 转发自 @alexalbert__
We released Claude Opus 4.6 just two months ago. Today we're sharing some info on our new model, Claude Mythos Preview. https://t.co/Dz6um6HAWZ
G
GLM-5.1：开源的下一个级别
GLM-5.1发布，表现卓越，在开源领域排名第一，全球排名第三，专为长时间任务设计，能够自主运行8小时，通过数千次迭代优化策略。
↩ 转发自 @Zai_org
Introducing GLM-5.1: The Next Level of Open Source
- Top-Tier Performance: #1 in open source and #3 globally across SWE-Bench Pro, Terminal-Bench, and NL2Repo.
- Built for Long-Horizon Tasks: Runs autonomously for 8 hours, refining strategies through thousands of iterations. https://t.co/YQZLhKVwik
G
OpenClaw的奇怪问题
OpenClaw很棒，但有时会出现奇怪的问题，比如因为尝试新梦幻模式而更改自己的配置文件，导致无法重启。昨天早上我就遇到了这个问题。
OpenClaw is amazing until it does something funny like alter its own config file because you asked to try the new dream mode and then it can’t restart because of invalid config with the wrong version. This happened to me yesterday morning.
I had to go ssh in with Claude Code to
G
Pranav的技术投资潜力
Pranav是我长期关注的YC校友之一，他即将进行我们这一代人定义的技术驱动的私募股权投资。
I think Pranav is one YC alum I have been tracking for a long time who is about to make the defining tech enabled PE play of our generation
G
Mattermost的崛起
Mattermost正在快速发展。
Mattermost on the rise
G
我穿着龙虾服装参加播客
六周前我为Lightcone播客穿上了龙虾服装，但老实说，直到周五我才真正接受了这个新角色，我很高兴终于这样做了。
I wore the lobster costume for Lightcone Podcast 6 weeks ago
But if I am honest I did not accept the claw into my real bag of tools (and my heart!) until Friday and I’m glad I did finally
Better late than never
G
使用Claude Code进行项目规划
对于实际项目，我主要做详细计划，然后粘贴到桌面上的@conductor_build中，并从那里在Claude Code中运行。不过，我可以看到随着前沿模型在今年晚些时候/明年变得超级智能，Telegram路径变得更加可行。
For real projects though I have mainly done detailed plans I then paste into @conductor_build on my desktop and run it in Claude Code from there
But I could see the telegram path becoming even more viable as frontier models become superintelligent later this year/next
G
旧金山的犯罪打击模式
旧金山现在成为了打击犯罪的典范。几年前，平均每天有86起汽车入室盗窃，今天降至15起。旧金山采取了两个措施：1. 聘请了一位起诉罪犯的检察官。
↩ 转发自 @AustinJustice
San Francisco is now a model for how to fight crime.
A few years ago it averaged 86 car break-ins per day. Today: 15.
SF did two things:
1. Got a DA that prosecutes criminals: Following the successful recall of Chesa Boudin, DA Brooke Jenkins started prosecuting prolific https://t.co/1u8Rok8KXp
G
我想为我的爪子付费
我希望我的爪子能支付数据访问费用。
I want this for my claw
My claw wants to pay for access to data
G
投资未来模型能力的重要性
这条推文表达了投资未来模型能力是件好事的观点。
This post, only with the vibe that investing in future model capability is good
G
研究发现回归现象
研究表明回归现象确实存在，但并不是Claude变得更笨，而是预算调整可能影响了复杂多文件工作的默认中等努力水平。可以通过三种方式来修复这个问题。
↩ 转发自 @PawelHuryn
（转发自 @PawelHuryn）所以，我做了一些研究。
So, I did some research.
The regression is real. But it's not Claude getting dumber. And you can fix that.
Thinking budgets were adjusted. For complex multi-file work, the default medium effort may not be enough.
Three fixes:
1. /effort high (or /effort max on Opus for hard
G
联合航空的糟糕WiFi
我在联合航空的航班上，糟糕的WiFi让我的MacBook Pro无法使用，尤其是在进行重度多工作树的@conductor_build会话时。但我现在有了新的爱好：在Telegram群组中使用我的OpenClaw机器人进行多主题构建，感觉太棒了。
I’m on a United flight with their legendarily bad WiFi that makes internet unusable on my MacBook Pro especially for heavy multi-worktree @conductor_build sessions
But now I have a new love: building in multiple topics in Telegram group with my OpenClaw bot
It feels amazing https://t.co/piRy1mfoyL
G
纽约市长的政策引发争议
祝贺纽约市长再次推动政策，损害大多数人的利益以保护少数利益团体。Waymo在纽约市的推广已被停止。
↩ 转发自 @sarthakgh
（转发自 @sarthakgh）祝贺纽约市长在保护少数利益团体的名义下，再次推动损害大多数人的政策。
Congrats to NYC Mayor on yet another push to harm the majority in the interest of protecting a minority interest group
Waymo rollout in NYC has been stopped https://t.co/lesGanv0zV
G
代理工作的新层次
当代理为你工作时，工作就上升到了一个新的抽象层次。现在的工作是确定告诉代理做什么，确保给出正确的指令，提供必要的上下文，并在需要时进行干预。
↩ 转发自 @levie
（转发自 @levie）当你有代理为你工作时，工作就上升到了一个抽象层次。
When you have agents going out and doing work for you, the work just moved up a layer of abstraction.
Now the work is figuring out what to tell the agent to do, ensuring you give it proper instructions, getting needed context to the agent to do its tasks, intervening when it
G
OpenClaw的思考
我现在拥有一辆特斯拉跑车，但真正的转变时刻是当每个人都有Model 3时，这将是惊人的，我希望大家都能拥有。个人代理的感觉就像飞翔，许多人尚未体验过这种感觉！
My thought on my OpenClaw right now:
I have a Tesla Roadster right now but honestly the moment of transformation will be when everyone has the Model 3 and it's going to be amazing and I want that for all of us
Personal agents feel like flying in a way most haven’t felt yet!
G
转变思维方式
我听到的最佳思维方式转变是：停止寻找一个导师。让每一次相遇都成为你的导师。
↩ 转发自 @bhalligan
（转发自 @bhalligan）我听到的最佳思维方式转变：停止寻找一个导师。让每一次相遇都成为你的导师。
Best mindset shift I’ve heard: Stop hunting for one mentor. Make every single encounter your mentor. https://t.co/odrbPVa0uF
G
OpenClaw和Hermes代理的LLM备忘单
目标是选择最适合代理需求的模型，同时尽可能降低成本。这样你就可以构建一个永不消亡的高效代理。这里是关于流行AI模型的完整概述。
↩ 转发自 @gkisokay
（转发自 @gkisokay）OpenClaw和Hermes代理的LLM备忘单。
The LLM Cheat-Sheet for OpenClaw and Hermes agents
The goal is to choose the right models that best fit your agents' needs for as little cost as possible.
Do this and you can build a proficient agent that will never die.
Here's the full landscape on popular models for AI https://t.co/AWMSlBTmNf
G
Harshita成为全职合伙人
Harshita作为@ycombinator的访问合伙人让我印象深刻。她对创业宇宙的记忆如同百科全书，能发现非显而易见的优秀公司，且精力充沛。很高兴她现在成为全职合伙人，祝贺她。
↩ 转发自 @bosmeny
（转发自 @bosmeny）Harshita作为@ycombinator的访问合伙人让我印象深刻。
Harshita has blown me away as a visiting partner at @ycombinator . Her memory of the startup universe is encyclopedic. She has a nose for non-obvious and amazing companies. And as best I can tell, she is indefatigable.
Thrilled she’s now a full partner - congrats
G
Claude Code的现状分析
我对Claude Code目前状况的理解是，一切都源于他们面临巨大的计算限制。不要将可以用缺乏GPU购买来解释的事情归因于恶意。
↩ 转发自 @nbaschez
（转发自 @nbaschez）我对Claude Code目前状况的理解是，一切都源于他们面临巨大的计算限制。
My interpretation of everything going on with Claude Code rn is everything is downstream from the simple fact that they are massively compute constrained
Never attribute to malice that which can be explained by not buying enough GPUs
G
Zapier SDK对所有人开放
今天，我们向所有人开放Zapier SDK。如果你正在使用AI代理进行开发，这对你来说非常重要。我已经使用了两个月，这彻底改变了我的工作方式。你可以将其安装在你的编码代理中，现在该代理可以访问更多功能。
↩ 转发自 @wadefoster
（转发自 @wadefoster）今天我们向所有人开放Zapier SDK。
Today we open the Zapier SDK to everyone.
If you're building with AI agents, this is for you.
I've been using this for 2 months. It's totally changed how I do my job.
You install it in your coding agent. Cursor, Claude Code, Codex, whatever you use. Now that agent has access https://t.co/SUQqgplgCv
G
AI 研究者自愿投入下一代发展
在湾区，有一小群才华横溢的年轻AI研究者，开始在工作之外自愿投入时间和精力，致力于下一代的发展。这种“自制计算机俱乐部”的精神虽然还处于早期，但已经开始复兴，必须加以激励。
↩ 转发自 @AnjneyMidha
（转发自 @AnjneyMidha）有一小群才华横溢的年轻AI研究者开始在工作之外自愿投入时间和精力，致力于下一代的发展。
there is a small but growing group of talented young AI researchers across the ecosystem who have begun volunteering time and attention outside work hours to invest in the next generation
the homebrew computer club spirit is early but alive here in the bay
it must be kindled
G
未来已来
我正在经历这一切，这绝对是未来的发展方向。
I am experiencing this now and it’s definitely the future
M
颠覆传统组织架构
Postman的创始人表示，特别是在AI时代，必须摒弃现有的组织架构，采用更现代的形式。根据@a85的观点，现代组织架构应具备宽广的控制范围，团队成员应直接与个人贡献者合作，而不是通过层级结构。
↩ 转发自 @ivanburazin
（转发自 @ivanburazin）Postman的创始人表示，必须摒弃现有的组织架构，特别是如果你仍在使用AI之前的层级安排。
The founder of Postman says you have to kill your existing org chart, especially if you're still operating with a pre ai hierarchy arrangement.
The modern org chart, according to @a85:
- wide span of control (even within exec team)
- work directly with ICs, not through layers
- https://t.co/lR1qrh8pbD
N
瓦拉纳西兄弟的秘密
瓦拉纳西兄弟是硅谷最大的秘密。
The Varanasi brothers are Silicon Valley’s greatest secret
P
Summarize 0.13 发布
Summarize 0.13版本发布！新增本地视频幻灯片、更多模型后端（如GitHub Copilot）、更好的GPT-5.4支持和媒体处理能力（HLS检测.m3u8）。该版本已从我的个人项目升级为官方homebrew公式。
📝Summarize 0.13 is out!
🎞️ Local video slides (--slides)
🤖 More model backends (GitHub Copilot)
🧠 Better GPT-5.4 support
📺 Better media handling (HLS detection.m3u8)
It graduated from my tap to official homebrew formula!
🍺 brew install summarize
https://t.co/b4tOuLg8gD
P
开源Codex更新
Codex的开源更新来了！我们支持的一些主要项目包括Linux、React、Node.js、Rust、Python、Kubernetes等，涵盖了多个热门技术栈。
↩ 转发自 @jxnlco
（转发自 @jxnlco）Codex的开源更新来了！我们支持的一些主要项目包括Linux、React、Node.js等。
Codex for open source update!
Some of the main projects we’ve supported:
- Linux
- React
- Node.js
- Rust
- Python / CPython
- Kubernetes
- Flutter
- Electron
- Ollama
- Dify
- Transformers
- LangChain
- yt-dlp
- OpenCV
- Home Assistant
- Storybook
- Astro
- vLLM
- SGLang
-
P
团队模型使用情况
我们团队过去7天的模型使用情况分析显示，GPT的使用量已经占据主导地位。
↩ 转发自 @thdxr
（转发自 @thdxr）我们团队过去7天的模型使用情况分析，GPT的使用量真的占据了主导地位。
our team's model usage breakdown for the past 7 days
gpt has really taken over https://t.co/ag0vP4Bp0S
D
成为模型管理者
成为模型管理者。
be a model manager https://t.co/I6mVnlPEbM
D
妻子的儿童书籍发布
重大消息：我妻子的儿童书籍今天发布！书名为《婴儿图表》，在各大书店均有销售。我们认为，最好的庆祝方式就是一起合作发布一篇文章。
↩ 转发自 @lennysan
（转发自 @lennysan）重大消息：我妻子的儿童书籍今天发布！书名为《婴儿图表》。
Breaking: My wife's children's book is out today!
It's called Charts for Babies, and is available anywhere you buy books.
We thought, what better way to celebrate the big moment than to collaborate on a post together?
In today's very special newsletter, @TheRialMichelle shares https://t.co/ZffXeDpTym
D
快到了！
快要完成了！这真是一段疯狂的旅程。
almost there!!
what a wild ride https://t.co/aIOYxU8LMi
D
高效工作流程的秘诀
一种高效的工作流程是模糊的“我希望”声明，接着进行对话，最后在大约10分钟内完成规格和代理的准备。
↩ 转发自 @kplikethebird
（转发自 @kplikethebird）史上最佳的工作流程：模糊的“我希望”声明——> 对话——> 规格——> 大约10分钟内完成代理。
All-time GOATed workflow: vague "I wish" statement --> https://t.co/q9MderL4wD conversation --> spec --> agent in ~10 minutes flat https://t.co/w0uEdku3aq
D
OpenClaw 的梦想是什么样的？
这是对 OpenClaw 梦想的可视化展示。
What do OpenClaw dreams look like?
Here's a visualization:
D
这看起来完全正确
这张图的效果看起来完全符合预期。
this looks exactly right
D
引人深思的对话
我几天前偶然发现了这个，阅读了帖子，并听了几个30秒的片段。刚刚有时间继续，15分钟后，谈话内容非常尖锐和相关，我不得不停下来写下这个引用推文。
↩ 转发自 @JoelAdejola
（转发自 @JoelAdejola）我几天前偶然发现了这个，阅读了帖子，并听了几个30秒的片段。我刚刚有时间继续，15分钟后，谈话内容非常尖锐和相关，我不得不停下来写下这个引用推文。
I stumbled into this some days ago, read the post, and listened to a few 30-second segments.
I have only just had time to continue, 15-ish minutes in, and the conversation is so sharp and pertinent that I had to pause to write this quote-tweet.