luolan_sun

AI产品3年，build in public，在 Startup 探索中

[luolan_sun @Luolan_AI]

@LerSentAI @fal 好的好的！

[luolan_sun @Luolan_AI]

@MaxForAI @internetphysics 啊啊这个好酷啊！

[Han Zheng @hanzheng_7]

Over the past few months, we’ve been thinking a lot about what it would actually take to build agents that continuously improve from their own experience. Today, we’re open-sourcing our continual learning infra, Reef.

The idea is simple: instead of treating inference as the end of the pipeline, Reef turns live agent interactions into a continuous learning loop. It serves real applications, captures trajectories and feedback as structured experience, and lets different learning recipes use that experience to improve the system.

What evolves isn’t just the model. Reef is designed to evolve the whole agent — model weights and the harness — then evaluate, version, and safely deploy those updates back into serving.

Really excited to finally share Reef we’ve been building toward continual self-improvement!

Come and check it out: https://t.co/IsAcJZGlFY

And join the Discord group for more updates: https://t.co/UcMW6CTJm4

[Paul Young @PaulYoungX]

Arrêtez de dire à Claude « fais ceci ».
Arrêtez de dire à Claude « écris du code ».
Arrêtez de dire à Claude « corrige cette erreur ».

Vous traitez en réalité une IA senior comme un stagiaire débutant.
Voici 8 prompts que vous pouvez copier-coller directement :

[luolan_sun @Luolan_AI]

@_WayWang @MeetResona cool！love this philosophy！Different people need different things from what you know.
wanna try it！

[Wei Wang @_WayWang]

What if your best lesson, sharpest pitch, or most trusted framework could serve 1,000 people at once—and still feel one-on-one?

That's what our team @MeetResona makes possible.

Resona is a professional live AI agent built for creators and educators, turning your existing materials into interactive live sessions.

With Resona, what you know finally earns what it's worth ⬇️
>Explain, answer, and adapt as users interrupt and ask questions in real time.
>Deliver personalized service at scale.
>Switch seamlessly between 36 languages.
>Set your own session duration and price, or offer it for free.

Meet Resona today.
Create your own live session for free.

[luolan_sun @Luolan_AI]

想求一个邀请码！

[luolan_sun @Luolan_AI]

既然知道一定会有结果，那就好好享受当下的过程。要么在等待中升级自己，要么在瓶颈期把身体养好，身体和脑子才是最好的投资。不要麻木，不要欲言又止，要时刻锋利，随时出击。

[luolan_sun @Luolan_AI]

我发现创业者应该多和创业者交流，用来��抗打工思维带来的怀疑和不安全感。高级打工人更在意title、package、stability，创业者更在意我选择解决哪个社会问题，我要怎么解决它？资源、判断力、执行速度。自己是否相信是致命一击。

[luolan_sun @Luolan_AI]

@0xwhrrari thanks！

[rari @0xwhrrari]

Google just released a free 2-hour course

On going from one prompt to a complete multi-agent graph:

17:44 - Build your first AI agent
39:30 - Create self-improving agent loops
1:12:38 - Connect those loops into graphs
1:34:26 - Make agents regulate themselves
1:55:05 - Orchestrate the full multi-agent system

Most people are still building one agent and stopping there

Google is teaching the entire stack:

Prompt → Agents → Loops → Graphs → Multi-Agent Systems

Single agents are the old workflow

Self-regulating agent graphs are the new one

This free 2-hour course is worth more than most paid agent engineering programs

Bookmark and watch it today

Then read the full architecture below ↓

[Zara Zhang @zarazhangrui]

Contrary to popular perception, most people will not adopt a new technology just because it makes them more efficient.

Instead, they’ll adopt new technology because:
- Someone they know who’s similar to them on some dimension adopted the technology, and got positive results (eg became cooler, richer, more popular/respected), and they wanna become like that person
- They feel like everyone around them has adopted the technology, and if they don’t, they’ll be left behind

Notice that it’s not entirely a rational decision. It’s more of an EMOTIONAL decision.

So if you’re trying to make someone adopt a new technology, the messaging shouldn’t be “this will make you 10x more efficient!” It should be “So-and-so, someone you can relate to, has adopted this and look how much better their life has become; wouldn’t it be nice to be like them?”

This book reminds you of that. It can be distilled in one sentence: “The diffusion of technology is a SOCIAL process”

[luolan_sun @Luolan_AI]

半夜发现Codex 更新了一个小功能，Subagent 在运作的时候，调用同一个智能体，会通过颜色标识，代表工作进程变化。

这让我对Codex的印象有所改观，之前总觉得Codex是理性克制的理工男，OpenAI的产品总显得很克制，不像Anastropic，有审美的、脾气。

他突然加上颜色标记，让我感觉理工男温柔了好多！

[luolan_sun @Luolan_AI]

为什么我的Claude code 今天又慢又笨！opus4.8 是因为发布会了sonnet5的原因吗？

[luolan_sun @Luolan_AI]

虽然Lenny忘记是谁说的了，但不重要，他说了一句更加警醒的话：最让你害怕的东西，其实就像一个指南针，指向你应该去做的事。

but just the thing that is scariest is,like,that’s a compass towards that’s what you should be doing.

[luolan_sun @Luolan_AI]

about fear

Anything that there is a fear,my advice is kind of lean in and ask:

okay,is there something I can do about it?

what is within my control?

Because sometimes the frustration comes from fear and feeling like,but everything is outside of my control

[luolan_sun @Luolan_AI]

Lenny 接着说：你害怕进入的洞穴里，藏着你追寻的宝藏。

The cave you fear contains the treasure you seek.

[luolan_sun @Luolan_AI]

fiona说，她很喜欢2句话，一是：如果你不害怕，你会做什么？what would you do if you weren’t afraid? 另一句话是：偶尔做点让你害怕的事。and do something scary once in a while.

Because that’s also usually how we grow.

[luolan_sun @Luolan_AI]

that was the one action I thought I could take that within my control to try to counter this fear I had of if I’m really gonna go down this path

[luolan_sun @Luolan_AI]

周一到周五上学，周六当收银员，赚学费。这就是多想想自己能干什么？