John Fan

🚀 Exploring AI commercialization, vertical applications & early-stage product testing. Sharing insights from the frontlines of intelligent innovation.

[John Fan @ainewmeth]

A fluent LLM is not a production worker.

A real enterprise task is a chain:
0.95²⁰ ≈ 0.36

That is why demos work, production breaks, and failures escape silently.

So I built agents like a Six Sigma system:
measure first, gate every step, turn failures into permanent controls.

Open source now:
https://t.co/Wa41SF9FwI

[John Fan @ainewmeth]

An LLM that's fluent in a chat box is not a worker you can put on a production line.

A real enterprise task is not one clever answer. It is a 20-step chain, and reliability is multiplied, not added:

0.95²⁰ ≈ 0.36

That is why the demo works, production breaks, and nobody can explain where the failure escaped.

Before AI, I spent ~13 years doing Six Sigma and process improvement on factory floors. So I built agents the same way we built reliable processes:

measure first,
gate every step,
separate execution from judgment,
and turn every real failure into a permanent control.

The result is open source now: a 7-skill suite for Claude Code, MIT licensed, distilled from my book.

One orchestrator runs the full pipeline:

assessment → control plane → guardrails → human review → independent measurement → DMAIC → production gate

https://t.co/Wa41SF9FwI

[John Fan @ainewmeth]

I've always said that philosophy is the ultimate major for the AI era.
In an age overflowing with answers, lacking the training to verify truth leads us to nihilism, bigotry, blind conformity, or cognitive atrophy.
But then again, tech giants are now aggressively poaching philosophy talents to train their models... so I guess our own ability to think deeply doesn't really matter anymore.

[John Fan @ainewmeth]

真嘟假嘟？

[John Fan @ainewmeth]

PHASE 4

[John Fan @ainewmeth]

承接FDE培训 免费分享

[John Fan @ainewmeth]

承接FDE培训 免费分享

[John Fan @ainewmeth]

最近 FDE 挺火，我理解这东西本质上是管理咨询+业务重构。我做了十几年的管理咨询，我们传统的项目里一个大难点是怎么用最快的方式找到最大的业务增长点。这得懂点商业。另一个大难点是怎么推动落地，变革管理，落地是最难的。只懂代码是不行的，光懂业务也不够，今年开始到现在，我就一直在做这件事。做诊断、画 Demo、搭Agent、做部署，这还得依托这些年积累的方法论啊、案例呀。更重要的是积累多年的客户信任，这些都具备了，都还挺难的。概念虽热但我感觉至少有 80% 以上的 FDE 很难真正交付

[John Fan @ainewmeth]

Totally feel this. Right now, most LLMs behave like that one insecure corporate yes-man who just wants to agree with the boss to avoid trouble. "You're totally right, what was I thinking!"A good human teammate has skin in the game; AI just wants to close the token ticket. I’ve found that telling the AI: "Your job is to be a critical partner, not a yes-man. Push back if I am wrong" helps a bit, but they still lack that genuine intellectual backbone

[John Fan @ainewmeth]

最近 FDE 挺火，我理解这东西本质上是管理咨询+业务重构。我做了十几年的管理咨询，我们传统的项目里一个大难点是怎么用最快的方式找到最大的业务增长点。这得懂点商业。另一个大难点是怎么推动落地，变革管理，落地是最难的。只懂代码是不行的，光懂业务也不够，今年开始到现在，我就一直在做这件事。做诊断、画 Demo、搭Agent、做部署，这还得依托这些年积累的方法论啊、案例呀。更重要的是积累多年的客户信任，这些都具备了，都还挺难的。概念虽热但我感觉至少有 80% 以上的 FDE 很难真正交付

[John Fan @ainewmeth]

@xiaohu 犹豫了好几天，想买个64G 的 Mac，这下错过。

[John Fan @ainewmeth]

@ZeroZ_JQ 这效果有点光污染，太多了，反倒没那么好看了。

[John Fan @ainewmeth]

https://t.co/Ba2cI6xGes
操，终于上线了个半成品

[John Fan @ainewmeth]

@james84_ Hello

[John Fan @ainewmeth]

随手一拍，image2.0 随手一修，完美

[John Fan @ainewmeth]

@idoubicc 这么好赚吗？

[John Fan @ainewmeth]

这样就可以了

[John Fan @ainewmeth]

coming soon

[John Fan @ainewmeth]

coming soon

[John Fan @ainewmeth]

@_daxiongya 可能不在企业，这种场景太少了

[John Fan @ainewmeth]

连公众号都满屏 Codex 的 Record and Replay，我实在想不出来这玩意有啥用，如果要消灭重复工作，那定时任务不都已经完成了吗？天天干重复工作，不应该马上被淘汰掉了吗？用 CodeX 的人真的需要这个吗？场景在哪里呀？大哥大姐们跟我讲讲。

[John Fan @ainewmeth]

happy

[John Fan @ainewmeth]

The key is authority never propagates along the chain — each stage gets a minimal scope granted at a gateway, not inherited from upstream, so a failed stage can't hand down permissions it never held. Add the lethal-trifecta split (no agent holds private data + untrusted input + outbound action at once) and fail-closed gates keyed to reversibility rather than confidence, and a single failure can't assemble an abuse path. It's a containment axis, orthogonal to the pⁿ reliability one.

[John Fan @ainewmeth]

An LLM that's fluent in a chat box is not a worker you can put on a production line.

A real enterprise task is not one clever answer. It is a 20-step chain, and reliability is multiplied, not added:

0.95²⁰ ≈ 0.36

That is why the demo works, production breaks, and nobody can explain where the failure escaped.

Before AI, I spent ~13 years doing Six Sigma and process improvement on factory floors. So I built agents the same way we built reliable processes:

measure first,
gate every step,
separate execution from judgment,
and turn every real failure into a permanent control.

The result is open source now: a 7-skill suite for Claude Code, MIT licensed, distilled from my book.

One orchestrator runs the full pipeline:

assessment → control plane → guardrails → human review → independent measurement → DMAIC → production gate

https://t.co/Wa41SF9FwI

[John Fan @ainewmeth]

Claude自己审自己确实不行，Codex+sigma-agent-engineering-suite，盘剥的干干净净

[John Fan @ainewmeth]

A fluent LLM is not a production worker.

A real enterprise task is a chain:
0.95²⁰ ≈ 0.36

That is why demos work, production breaks, and failures escape silently.

So I built agents like a Six Sigma system:
measure first, gate every step, turn failures into permanent controls.

Open source now:
https://t.co/Wa41SF9FwI

[John Fan @ainewmeth]

@fankaishuoai 实在是没想出来这破玩意有啥用，说句实在话，周报也写不了啊