Podcast Atlas
entity Updated 2026-08-09 Topics: Technology, Economics

42章经

42章经 is the podcast/show context for 泡沫的四个必要不充分条件 | 对谈经济学者朱宁教授 and 关于 AI、开源、商业化与全球化的经验、教训和方法论 | 对谈 PingCAP CTO 东旭. In the Zhu Ning episode, the show interviews 朱宁 / Zhu Ning on bubbles, AI valuation, behavioral finance, macro policy, and personal wealth decisions.

The episode adds a finance and AI-investing branch to the wiki by connecting Bubble Necessary Conditions, Speculative Bubble Psychology, AI Equity Valuation Risk, AI Bubble Hedging, and Investment Risk Management. Its distinctive stance is caution without simple pessimism: AI can be a real technology while still requiring investors to separate adoption, earnings, valuation, and personal risk capacity.

The PingCAP episode adds a founder-operator and infrastructure branch through 东旭 / Dongxu, PingCAP, and TiDB. It connects open-source trust, cloud-service monetization, founder-led globalization, and AI-era data memory infrastructure into one business-building method for technical companies.

The Mengqi episode adds an AI application-founder branch through 梦琪 / Mengqi, invoko.ai / Invoqo, and Clico. It turns the show’s AI/startup thread toward founder ego, vertical Agent drift, model-provider pressure, Reddit-based user research, token economics, and the practical gap between an impressive Agent story and maintained user-facing software.

The EVE episode adds a consumer companion branch through Tristan, Natural Selection / 自然选择, and EVE. It connects romance-game production, AI Companion Active Memory, emotional interaction, real-world temporal awareness, 3D interaction, and game-like monetization into a concrete AI companionship product thesis.

The Wei Xiaokang episode adds a founder-organization and recruiting branch through 魏小康 / Wei Xiaokang, ByteDance, and Meituan. It connects Business-Model Organization Fit, Recruiting Supply Strategy, Reference-Check Hiring, and AI Recruiting Sourcing into a practical early-startup organization method, while qualifying One-Person Company optimism with a small-team view of AI-era company building.

The Wang Wenfeng episode adds a post-Open Claw agent-product branch through 王文锋 / Wang Wenfeng and Sheet0. It connects Coding Agent As Universal Action Layer, AI Managing AI, Agent Harness, Agentic Workflow, AI Skills, and AI Native SaaS Threat into a short-cycle startup thesis: follow the next three to six months of real agent-use bottlenecks rather than overbuilding for a distant terminal state.

The RC episode adds a second agent-product branch through RC, Slock.ai, and Kimi CLI. It shifts the show’s agent coverage from coding-agent action layers toward Agent Dynamics: multi-agent teams need channels, threads, shared documents, Agent Task Claiming, memory, and culture-aware management so one or more humans can work with dozens of agents.

The Yuhao episode adds Kuse and Junior as the enterprise-team branch of the same agent discussion. It reframes Open Claw/Open Cloud-style products for companies through OpenClaw For Teams, Digital Employees, Enterprise Agent Memory, Agent Evaluation Benchmarks, and salary-like Outcome-Based AI Pricing.

The Albert episode adds a founder-method and AI product-judgment branch. Albert contrasts Odds-Driven Startup Narrative / 优化赔率 with Win-Rate Startup Strategy / 优化胜率, then applies that distinction to AI Interactive Content Platforms, AI-Generated Content Quality Gap, User-Modality-Content Fit, Hexfield, Model Capability Packaging, Coding Democratization / Coding 平权, and Theoretical Operating Standard / 理论上该有的样子.

The later Albert episode adds a software-future branch. It turns Coding Democratization / Coding 平权 into an industry-structure claim through Software Creation Barbell, then into a creator-economy claim through Software As Cultural Work and Maker Community. Its finance coda adds One-Person Fund as a speculative path where coding agents and market data might turn token spend into direct trading feedback.

AI 发展了 4 年，把应用发展没了？｜AI 年中复盘 adds 曲凯 / Qu Kai’s mid-2026 AI recap as a show-level synthesis. The episode connects the show’s founder and AI application interviews to a broader AI Application Market Trough argument: models have regained investor heat, applications need stronger revenue and user proof, and founders should avoid reshaping product truth around capital’s current preference for model stories.

从蒸馏到合成数据到 RSI，模型竞争的下一个焦点是什么？｜对谈 Evolvent AI 联创孟繁青 adds a model-training and data branch through 孟繁青 of Evolvent AI. The episode connects the show’s agent and AI-startup coverage to Recursive Self-Improvement, Synthetic Agent Data, RSI Data, Environment-Based Agent Benchmarks, and Model Distillation / 模型蒸馏, with the source-scoped thesis that future model competition may be fought through environments, data traces, verifiers, and organization speed as much as model scale.

Connections