@laowangbabababa: Hilarious. Big tech companies like Alibaba and ByteDance never expected their own jargon would be used to PUA AI. 18.8k stars, open source, MIT. The project is called PUA AI, packed with 14 types of big company lingo, automatically switched based on the task. If you ask AI what the underlying logic is, it goes into Alibaba closed-loop mode. If you ask AI about ROI, it switches to ByteDance data and A/B testing…

X AI KOLs Timeline Tools
open-source ai-efficiency prompt-engineering chinese-tech pua-ai productivity-tool experiment

Summary

The project PUA AI collects 14 types of internal jargon from major Chinese tech companies and automatically switches between them to guide AI behavior. Experiments show it improves fix points by 36%, verification steps by 65%, and tool invocation and hidden issue discovery by 50%.

Hilarious. Big tech companies like Alibaba and ByteDance never expected their own jargon would be used to PUA AI. 18.8k stars, open source, MIT. The project is called PUA AI, packed with 14 types of big company lingo, automatically switched based on the task. If you ask AI what the underlying logic is, it goes into Alibaba closed-loop mode. If you ask AI about ROI, it switches to ByteDance data and A/B testing. In other words, it's like equipping every agent with a full set of corporate methodologies, for free. It also sets three hard rules that can't be faked. - Don't give up until you've tried everything. - Must use available tools first. - After finishing, verify and also check similar issues. Previously, Old Wang's agent would claim "I can't solve this" after running the same command three times. Not anymore. It's not just a toy — it really improves AI efficiency. 18 sets of controlled experiments show 36% more fix points, 65% more verification steps, and 50% more tool calls and hidden issues discovered. It doesn't just feel stronger; it actually does more work. Big companies used to pressure people with this. Now developers turn around and stuff these scripts into AI, really cranking up the pressure on AI. Repo link in the comments. P.S. The first line of the repo description made me laugh — "You are a highly anticipated P8-level engineer. When Anthropic originally rated you, their expectations were sky-high."
Original Article
View Cached Full Text

Cached at: 07/20/26, 09:33 PM

This is hilarious. Big tech firms like Alibaba and ByteDance probably never imagined their own corporate jargon would be used to PUA AI.

18.8k Stars, open source, MIT. The project is called PUA AI, packed with 14 types of big tech lingo, automatically switching based on the task.

Ask AI what the underlying logic is, and it goes into Alibaba’s closed loop. Ask about ROI, and it jumps into ByteDance’s data and A/B testing.

Thinking about it differently, it’s essentially equipping every agent with a full enterprise methodology — for free.

It also sets three hard rules that can’t be bypassed:

  • Don’t give up until all options are tried.
  • If you have a tool available, you must use it.
  • Once done, verify and also check similar problems.

Old Wang’s agent used to run the same command three times and then claim “I can’t solve it.” Not anymore. And it’s not just a toy — it actually boosts AI efficiency.

18 controlled experiments: 36% more fix points, 65% more verification steps, 50% more tool usage and hidden issue discovery. It’s not just feeling stronger — it’s actually doing more work.

Big tech used to use this playbook to drive people. Now developers are turning around and feeding that jargon into AI — for real.

Repo link in the comments.

P.S. The first line of the repo description made me laugh — “You are a P8-level engineer who was once held to high expectations.” Anthropic had high hopes for you when they gave you that rank back then.

Similar Articles

@yanhua1010: I never had high expectations for domestic AI Agents, until this time it did something that surprised me. I asked Tencent WorkBuddy to do a competitive analysis, and it dispatched an entire research team, with one AI specifically for review. It sent back the report written by its colleagues because a number in it had no verifiable source...

X AI KOLs Timeline

A user shares their experience using Tencent WorkBuddy for competitive research. The AI deployed a research team, and a dedicated review AI caught a number without a source, rejected the report, and had it redone. The final output was a presentation-ready slide deck.

@ai_laotie: https://x.com/ai_laotie/status/2068215413050347654

X AI KOLs Timeline

Introduces three AI workflows that will make money in 2026: reverse-engineering overseas case studies, converting long-form articles into viral X threads, and building a faceless AI short video matrix. Emphasizes that systematic workflows and prompt iteration matter more than chasing new tools.

@ba_niu80557: Let's talk some hardcore practical knowledge while I have time this morning. What actually happens between signing a contract for an AI project and it finally running in production? I'll lay out the entire playbook. People in this field can copy it directly, and those not in it can still understand why 95% of enterprise AI pilots end up dead. First, let me say something counterintuitive to the point you might not believe...

X AI KOLs Timeline

This article discusses common reasons for the failure of enterprise AI projects from proof-of-concept to production deployment, highlighting key practices such as MLOps, early inspection of real data, and clear human-machine boundaries. It argues that project failures are often not due to model issues but due to neglect of the engineering implementation phase.

@blueskylh1: The most painful thing about solo product development or leading an AI team is being a "mindless messenger" between different chat windows. After the PM writes the requirements, I have to copy and paste them into the developer's chat. After seeing the sharing from Jason @jxnlco, a developer experience engineer on the OpenAI Codex team, I set up a workflow without...

X AI KOLs Timeline

Introduces a multi-AI agent collaborative workflow based on local plain text files and OpenAI Codex, allowing PM, backend, frontend, and QA to efficiently develop via file relay without copy-pasting.