Quan Xu｜AI GEO/GEM

Founder of All Intent AI｜11-year B2B AI marketing veteran, back in the arena for the age of AI-powered global growth｜Helping Chinese business owners win custome

[Quan Xu｜AI GEO/GEM @quangeAI]

@tangjinzhou 唐老师的X粉丝更稀缺，哈哈哈

[Quan Xu｜AI GEO/GEM @quangeAI]

不入局，是实现零消耗的最高智慧

[Quan Xu｜AI GEO/GEM @quangeAI]

围棋，比象棋高明的地方，围棋是大局为重，象棋是厮杀

[Quan Xu｜AI GEO/GEM @quangeAI]

人越被信任，越被允许，越被理解，越被爱，就越强大。信任和爱会滋生力量

[Quan Xu｜AI GEO/GEM @quangeAI]

@Banter_Plane 是的，但我发现越来越多人已经看不出来了

[Quan Xu｜AI GEO/GEM @quangeAI]

很多人以为用上codex就天下无敌，但还是要有算法思维才行。如果没有的话，codex犯傻的时候，人也跟着一起傻，因为根本看不出来

[Quan Xu｜AI GEO/GEM @quangeAI]

#写作
AI时代，按理说写作就可以完全AI来替代，我也知道很多人的内容都是AI写的，但是我现���有个反共识的认知，AI时代反而要坚持✍️手写，因为文���的力量，在写作的时候会注入情感和能量，这才是打动人的基础。AI写出来的东西，道理都对，但就是没有灵魂。写作也是一种思维训练对我来说

[Quan Xu｜AI GEO/GEM @quangeAI]

#教育
有人真的活出来个性化教育的样子。在当下的教育体制下。
结果查了下这个石室中学历史，就可以理解了。
​
​最好的教育，不是把所有人塑造成同一个模样，
​而是守护每一个盖世英雄的梦想。

[Quan Xu｜AI GEO/GEM @quangeAI]

见证苹果新时代，作为同样从硬件出身又做了软件出身，我还是挺看好特纳斯可以带领苹果重回乔布斯时代的创新力，现在运营是很强，但苹果真正的核心在创新

[John Ternus @johnternus]

hello

[Quan Xu｜AI GEO/GEM @quangeAI]

人与人的差别，
​除了先天因素，
​剩下来的就是心力和思维结构的不同了。

[Quan Xu｜AI GEO/GEM @quangeAI]

@skaas777 哈��哈，牛逼了

[Quan Xu｜AI GEO/GEM @quangeAI]

开源精神常在，但是国内的OpenClaw的风已经被腾讯的workbuddy，字节的豆包工作，阿里的千问办公抢了。只能说，AI的风，来得快，去的也快。什么才是不变的。

[OpenClaw🦞 @openclaw]

OpenClaw 2.0 has arrived
https://t.co/52iWXJQvCy

[Quan Xu｜AI GEO/GEM @quangeAI]

我们总在回望中读懂一个时代，
身处时代洪流之中，却往往浑然不觉。
AI这可能是人类历史上最后一次技术革命了，我们一起做出一些有意义的事情，不辜负这个伟大的时代。

[Quan Xu｜AI GEO/GEM @quangeAI]

有时候跟人协作，效率真的太慢了，还不如自己来，自己来又没有组织的规模化效应，人就是在不断矛盾的阶段前行的

[Quan Xu｜AI GEO/GEM @quangeAI]

@JimBlade 哈哈哈，没你注册得早呀

[Quan Xu｜AI GEO/GEM @quangeAI]

去年的一个聘书，#AI
​#乡村振兴
​感谢全国大学乡村振兴创意大赛的认可，
​AI和乡村振兴会碰撞出什么火花呢？
​期待为祖国的下一代，更年轻的年轻人。
​帮他们少走弯路，进入AI的世界。
​AI不是战术，AI是战略。
​AI不是工具，AI是时代。

[Quan Xu｜AI GEO/GEM @quangeAI]

the underlying attention deficit of LLM is precisely the core entry point of AI GTM, better credibility weight, and ultimately enter the pre-training library of LLM.

[Quan Xu｜AI GEO/GEM @quangeAI]

谷歌的论文向来很强，研究底层技术值得研究，但是就是最近谷歌的gemini模型有点销声匿迹了。

[Akshay 🚀 @akshay_pachaar]

Google just dropped a banger paper.

If you write or curate skills for your agents, this one is for you.

(bookmark it)

Agent skills are just folders holding the instructions an agent follows for a task, and people increasingly let agents write those instructions themselves.

The loop is simple. Run the agent on some tasks, read the runs that failed, rewrite the instructions, and keep the rewrite only if the score goes up on a held-out validation set.

That last check matters, because an edit that quietly makes the agent worse would pile up over time. So the system always applies the edit, tests it, and reverts the file if the score drops.

Here is where it breaks. Working out why the agent failed is the expensive part, since it means reading full traces and comparing failed runs against successful ones.

But that diagnosis never gets written down. The only thing saved is the new instruction text it produced.

So when that instruction reverts, the finding behind it goes too. The next round reads the same failures, reaches the same conclusion, and often proposes the fix that already lost, with no memory that it was ever tried.

WikiSkill fixes this by splitting the one folder into three:

1. Raw traces, written once and never touched.

2. A wiki of accumulated knowledge, holding what keeps breaking, what has worked, and every edit tried so far with the reason it passed or failed. Nothing here is ever deleted.

3. The skills themselves, reverted whenever an edit makes things worse.

So the instructions can roll back, but the understanding behind them never does.

Their own example makes it click. An early skill gets rejected for being too vague, and instead of vanishing, that rejection is logged.

The next round reads the log, sees the vague version failed, and writes a concrete rule in its place. That one gets accepted.

Same failure, same analysis, but the second attempt starts from knowing what already did not work.

The authors confirm this is where most of the gain comes from by turning the knowledge layer off. The average score drops by about a quarter.

Two other results are worth the read on their own.

Smaller models with evolved skills beat much larger models running without any, and skills evolved by one model transfer across families, sometimes working better than the ones a model wrote for itself.

Paper → https://t.co/ungqaQVTma

I wrote about a related idea (GEPA) earlier, which improves a model by having it reflect on its own mistakes in plain language and rewrite its prompt, with no weight updates involved.

The detailed article is quoted below.

[Quan Xu｜AI GEO/GEM @quangeAI]

#阅读
​书中自有黄金屋
​书旁自有小猫咪[破涕为笑]

[Quan Xu｜AI GEO/GEM @quangeAI]

现在更瘦了，创业需要有个好身体，目标到140的标准体重