Podcast Atlas

Updated · 12 episodes · 2 shows · 12 source notes

entity

刘飞

Overview

刘飞 is a Chinese podcast host and former internet product manager connected to 半拿铁 and 三五环.

Current Profile

The current wiki profile presents Liu as a product-minded interviewer whose podcast work spans business storytelling, media-form reflection, AI-era creator questions, platform operations, founder workflow, AI-native IP, commerce rights, platform marketing science, and technology-community history. In the 半拿铁 special, he helps explain podcast production, audience trust, and brand communication. In the 三五环 interviews with 吴瀚坤, 抽象仔 / 抽象宅, and 汗青, he moves that product lens into AI filmmaking, actor displacement anxiety, prompt communication, internet expression, traffic, virtual-idol personality systems, and the question of what creators want to express when tools lower execution barriers.

The Didi safety episode adds a different version of the same interviewer profile. Liu uses prior coworker context and product intuition to draw 徐晓楠 / 曲晓楠 from abstract platform safety into concrete workflows: where a ride’s safety boundary begins and ends, why late-night phone calls happen, what AI can and cannot infer, and how users affect the platform’s ability to intervene.

The AHA Creator episode adds Liu’s founder-operator questioning around AI commercialization. He frames the conversation around whether AI is solving real productivity problems, then pushes Kate to explain overseas creator marketing as a concrete chain of search, matching, communication, negotiation, payment, review, and customer approval.

The Yuri / 尤栗 / 优瑞 episode adds Liu’s AI-native content and original-IP questioning. He draws 汗青 from AI Talk’s viral public-figure dialogue experiments into the harder questions of why an AI idol can move people, when realism becomes collapse risk, and why personality, narrative, and judgment remain scarce after model output becomes abundant.

The 《女娲之死》 episode adds Liu’s AI-filmmaking workflow questioning with 叫兽易小星 and 李晨. He uses the B站 AI 创作大赛 context to draw out how AI short films are actually made: script, storyboards, image generation, model choice, editing, finishing, hallucination checking, and responsibility after output.

The 抖音电商 infringement episode adds Liu’s platform-governance questioning around AI impersonation. He uses concrete live-room, celebrity voice, product-detail, account, and authorization examples to pull 朱威 and 嘉旭 into the practical boundary between AI as a commerce tool and AI as rights misuse or consumer deception.

The Kuaishou seeding episode adds Liu’s marketing-science questioning. With 刘晓微, he pushes from the surface language of “种草” into user paths: whether search, store browsing, live-room explanation, off-platform purchase, delayed conversion, and podcast advertising can be understood as positive signals inside 品效合一 rather than only as failed immediate ROI.

The 宇宙 E-Bike episode adds Liu’s founder-operator and manufacturing questioning. With 程泓宁 / 陈鸿宁, he frames the move from investing to entrepreneurship through operating control, selling a high-ticket hardware product, category education, China-Europe e-bike differences, and whether a small manufacturing company can become AI-native.

The 少数派 episode adds Liu’s community-history questioning. With 老麦 / Michael, he reconstructs the path from Palm and webOS tinkering through 主机网, app recommendation, author compensation, paid courses, product co-creation, and AI-era content scarcity.

The 《第二天金钱进化论》 episode adds Liu’s wealth-and-life-design interview branch. With 方言, he frames “having money” as a second-stage problem involving luck, lifestyle inflation, business overreach, family strain, and whether assets can return time rather than become public identity.

Key Characteristics

  • He connects internet product experience with long-form podcast interviewing.
  • He is associated with both 半拿铁 and 三五环 in the wiki’s source set.
  • He frames podcasts as relationship and trust media rather than only information delivery.
  • He uses product and workflow questions to make AI’s effect on creators concrete.
  • He often turns technical change back toward human purpose, audience expectation, and creator agency.
  • He can translate AI-enabled rights infringement into user-facing questions about consumer trust, authorization, and platform responsibility.
  • He can translate platform safety, AI commercialization, virtual-idol risk, marketing science, original-IP work, hardware entrepreneurship, community history, and wealth-after-success risk into user-facing questions about trust, workflow, demand, responsibility, and life design.

Evidence

Qualifications

The wiki profile is source-limited to podcast appearances and does not attempt a full biography. Liu Fei is distinct from 刘肥, the historical Qi king page with a semantic suffix.

What Changed

  • Added the No.220 Fang Yan episode, extending Liu’s profile from AI/product practitioner interviews into wealth-after-success, re-poverty risk, and life-design questioning.

Relationships

Sources

12 source notes across 2 shows
  1. 番外 14：跟李诞聊聊播客、创作、AI 与中年 半拿铁 | 商业沉浮录
  2. No.232 吴汉坤：AI 时代，演员第一个丢饭碗？ 三五环
  3. No.231 抽象仔：从《航拍中国》到《新鸳鸯蝴蝶梦》，重新用 AI 学习互联网表达 三五环
  4. No.228 对话滴滴曲晓楠：怕你觉得我们不安全，更怕你觉得我们绝对安全 三五环
  5. No.226 AI 越强，创业者越要回到用户现场 三五环
  6. No.225 对话 Yuri 尤栗打造者汗青：AI 虚拟偶像，越像真人越容易塌房？ 三五环
  7. No.224 用 AI「偷」明星声音和形象侵权卖货，直播间该怎么管？ 三五环
  8. No.223 当单纯的投放逻辑已经失效，我们如何重新理解种草？ 三五环
  9. No.222 程泓宁：从庙堂之上的投资人，到在真实的泥地里打滚 三五环
  10. No.229 对话少数派老麦：从 Palm、WebOS 到 AI 硬件，依然不想成为多数派 三五环
  11. No.227 对话叫兽易小星&粉墨：人人都能用 AI 拍电影了？ 三五环
  12. No.220 钱为什么不是万能良药？聊聊那些返贫的故事 三五环