Teigen

🤖 大厂AI算法工程师
🎓 人大计算机本硕
🌱 ENFJ｜认真生活 持续思考
📚 分享AI实践｜效率提升｜自我探索

[Teigen @Teigen27]

从1月份开始，基本每天都会用语音记日记。今天在推进美好时光日志的部分，直接使用Claude Code对过往的日记进行分析，提炼出了很多有趣的东西，记录的价值在此刻浮现。
​
​之前开源出来的基于《人生设计课》的项目，有陆续收到一些令人高兴的反馈，但也有朋友因为使用Claude Code存在一定���槛没能用上。今天在文档里面补充了国内免登录使用Claude Code的方式（P4，详见文末Github的README），希望能帮助更多人从中获益。
​
​就算没用上这个练习，光是能使用Claude Code，就足以给未来带来很多可能性。

介绍网站：https://t.co/y078TppsDp
GitHub：https://t.co/XtVkgxmFLp

[小互 @xiaohu]

Claude code 发布 多任务统一窗口管理工具：Agent View

以前你同时跑多个 Claude Code 任务，要开一堆终端窗口、标签页，分不清哪个任务在等你确认、哪个还在跑、哪个已经做完。

现在 Agent View 把会话集中到一个界面里管理。

你可以一眼看到所有Agent 的状态， 谁在工作，谁等你输入，谁已经结束。

它现在很像：

AI 多任务调度中心

你可以：

• 同时看到所有 Claude Code 窗口状���
• 任意切换到某个 Agent 对话
• 某个 Agent 卡住时立刻处理
• 某个 Agent 完��时立刻接管
• 不需要再回原 terminal

这个思路其实已经很接近：

“AI IDE 的任务栏”

[Thinking Machines @thinkymachines]

People talk, listen, watch, think, and collaborate at the same time, in real time. We've designed an AI that works with people the same way.

We share our approach, early results, and a quick look at our model in action.

https://t.co/AFJZ5kH7Ku

[Peter Steinberger 🦞 @steipete]

Whenever I investigate a bug, I let codex recreate the exact state in an emphemeral crabbox, verify the bug, fix it, verify the fix.

No messy state because local system might be polluted, and no slowdown because I run 10 sessions in parallel. https://t.co/SEj2XRpaD1

[Teigen @Teigen27]

最近的状态变化很大，vibe coding时间少了很多，看极其冗余的token额度也不焦虑了

没有那种被token推着走的感觉，似乎也就能承载住足量的额度，使其成为“资源”而不是“诅咒”
​
​时间更多流向了和真实的人的接触、向内的探索、情感的流动，AI在生活中存在的方式柔和了很多，生活的根是更扎实的，我会更喜欢现在的状态
​
​大概从3月份就在开发一���自我探索的AI产品，但迟迟未能内测，一直在优化，似乎没有尽头。最近才意识到是这个产品的根还没扎好，又在重新从源头去塑造画面感。AI是放大器，根足够稳固，放大后才不会倒塌

[Xin Eric Wang ✈️ CVPR 2026 @xwang_lk]

The bitter lesson of agentic memory: any human-curated memory module is fundamentally limited. Don't design memory (even agentic memory). Let agents evolve it under the pressure of doing the task better.

[Simon Willison @simonw]

Some notes on talkie, a new "vintage language model" from a team including Alec Radford (yes, that Alec Radford) "trained on 260B tokens of historical pre-1931 English text" https://t.co/BSFHG7ewg7

[Ethan Mollick @emollick]

An easy way to get a team engaged with AI is just to build the thing you are talking about in the meeting during the meeting using Codex or Claude Code.

At worst, it fails in ways that can be constructive. At best, you built the thing and the meeting topic shifts forward a month

[Teigen @Teigen27]

上周末跟朋友们聊到 AI 工具推荐，简单来说有两个观点：
1. 如果想深入了解 AI 并���有一定经济基础，推荐尽量使用付费的 AI 模型
2. 在当下这个时间点，如果仅推荐一个会员订阅，我会推荐 ChatGPT 100 刀的 Pro 订阅

一、关于推荐付费模型的原因
因为现在 AI 的飞轮转得越来越快，最强大的模型和没那么好的免费模型间的差异会越来越大，这一点可以参考 Karpathy 最近的推 https://t.co/yp9VBi2mLb

二、关于推荐 ChatGPT Pro 订阅的原因
1. 记忆系统：ChatGPT 的记忆系统做得非常好，这对日常聊天影响很大，Pro 订阅的记忆系统相比免费版会有强化
2. 模型能力：GPT-5.5 发布之后，综合能力非常强，并且 Pro 档位可以用 Pro 的模型去做非常复杂的任务
3. Codex 体验：订阅后可以用 Codex，Codex 的 APP 会比终端体验要更好，这对大部分人来说更容易上手。CC 源���泄露之后，Codex 近期的体验也有明显提升，现在已经可以跟 CC 掰手腕了
4. 养龙虾：订阅账户的额度可以用来养龙虾（OpenClaw），这意味着省下了养龙虾所需要的 API 费用
5. 额度充足：现在到 5 月底是这个档位的促销期，最多能够使用 10 倍的 Plus 额度，而且最近经常重置，额度很足
6. 风控宽松：ChatGPT 的账号风控没有 Claude 那么严格，没那么容易封号

三、个人使用习惯
如果还想再订阅其他家的 AI，我会推荐 Claude 的 Max 档位（100 刀或 200 刀，根据实际需求确定）
我个人一般是：
(a) 对话类任务：用 ChatGPT
(b) 代码类任务：用 Claude Code 搭配 Codex（在CC里面安装 Codex 插件）
(c) 养龙虾：直接用 Codex 额度

四、支付方式
1. 安卓：可以使用 Google Play 加上国内的全币种信用卡
2. ���果：可以去支付宝买礼品卡，然后通过美区 App Store 账号订阅

[Teigen @Teigen27]

在看《人生设计课》的时候，觉得里面的练习很有价值，但好像很难真正做起来，于是弄了个互动式练习项目，并开源出来了

底层逻辑是「寻路思维」—— 不问 “你想成为什么”，而是问 “什么让你活过来”

介绍网站：https://t.co/ZjJXKNHNfg
GitHub：https://t.co/DCIwGYeOBT

[Teigen @Teigen27]

把我的龙虾每天生成的简报给开放出来了，包括各个高质量AI信源(Youtube、Podcast、Blog、Techmeme)、X推文以及即刻动态。每日19点自动更新，通过网站即可访问 https://t.co/JnQ3lJVUYu

也同步提供了RSS订阅链接（网页右上角可获取），可以导入到RSS订阅器中

[Andrej Karpathy @karpathy]

Judging by my tl there is a growing gap in understanding of AI capability.

The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code.

But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along.

So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions.

TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

[Teigen @Teigen27]

Claude Code 现在可以直接调用 Codex，搭配起来非常舒服

Codex 找 bug 能力体感很强，能够在一定程度上补充 Opus 的盲点

对于复杂任务，可以开启���束前强制 Codex 审查，直到找不到问题再完成执行，感觉是一种更高效的 Ralph Loop

插件链接：https://t.co/75zKIuTmqx

[Ivan Zhao @ivanhzhao]

https://t.co/fxHYR6S29D

[Teigen @Teigen27]

原链接：https://t.co/I67EBgTMFH

[Pascal Bornet @pascal_bornet]

Google just showed the most beautiful use of AI I’ve seen this year.

No glossy ads.
No corporate jargon.
Just a simple idea — letting their employees’ children imagine.

They took a child’s hand-drawn sketch and used Veo, Imagen, and Gemini to turn it into a cinematic story.
And suddenly, it wasn’t about products or features anymore.
It was about imagination meeting technology.

I think that’s the real magic of AI.
It’s not replacing creativity — it’s amplifying it.
It helps people turn what’s in their head into something real, visible, shareable.

For centuries, we’ve told stories through whatever tools we had: stones, brushes, cameras.
Now, we have intelligence itself as a tool.

To me, this moment reminds us that AI isn’t just about what machines can do — it’s about what humans can dream again.

But it raises a question I can’t stop thinking about:
Will AI help us rediscover our creativity, or drown it in noise?

#AI #Creativity #Storytelling #Innovation #Technology #Leadership #FutureOfWork #HumanIntelligence

[Teigen @Teigen27]

将手绘草图，变成电��般的故事

[WaveSpeedAI @wavespeed_ai]

From a plain map to a full cityscape in one move.🪐
Streets, blocks, buildings — everything pops into place instantly.
World-building has never been this fast, this smooth, or this fun.
Ready to create yours?😜

[Teigen @Teigen27]

原链接：https://t.co/u9bBTazEPI

[Nuri Yıldız @nouryyildiz]

Hollywood Selfie Part ✌🏻

[Teigen @Teigen27]

AI视频：好莱坞自拍

[Teigen @Teigen27]

原链接：https://t.co/jaClU6VqLt

[Burak Tuyan @buraktuyan]

Thrilled to share “This Too Shall Pass”, my creative take on “Save the Christmas” for the Wondercraft Christmas Challenge.

Thank you @Alex_wonders_ and @dimiwonders for the invitation and for being there for the contestants throughout the process.

Created using tools inside @wondercraft_ai, in particular:
- Nano Banana Pro
- Kling 2.6
- Kling O1
- ElevenLabs Music
- ElevenLabs SFX

[Teigen @Teigen27]

AI圣诞视频：这终将过去