E45 孟岩对话李继刚：人何以自处
Summary
This 无人知晓 episode has Meng Yan / 孟岩 and Li Jigang / 李继刚 use investing, reading, prompting, company form, education, and existential anxiety to ask how people should live under AI. Li’s core move is to treat AI as AI As Time Compression: the internet removed much of the spatial cost of access, while AI compresses the time needed to search, read, synthesize, and receive feedback. The answer is not to surrender the person to models, but to cultivate Wet-State Human Agency, Feed Curation, Prompt As Intent Transmission, Personal Knowledge Ecology, and Water And Fire Education so the human keeps intention, taste, body, connection, and judgment.
Key Claims
- Li’s “frame” and “order” vocabulary turns reading and investing into a search for the small number of variables that shape a possible world, not a chase after opinions or price motion.
- AI As Time Compression explains why AI feels faster than the internet: a person can move from downloading books to asking a trained “crystal” of knowledge for structured dialogue.
- The episode’s human split is body power, brain power, and heart power; as AI absorbs more brain work, Wet-State Human Agency becomes the human layer around desire, emotion, will, aesthetics, and embodied connection.
- AI Company Deep Well contrasts internet-era networks with AI-era wells: an AI product may create value by knowing a smaller number of users much more deeply through context, memory, and value alignment.
- Trust As Business Asset becomes fragile in AI recommendation because inserting ads into a high-trust assistant relationship can damage the user’s belief that the answer is for them.
- Prompt As Intent Transmission treats prompts broadly: a prompt can be a role, uploaded document, local note, memory file, context folder, or any medium that carries a person’s will into the model.
- AMV Prompt Framework breaks prompting into A, the starting role or position; V, the destination or direction; and M, the desired shape of thought between them.
- Persistent Agent Memory and “Soul” are presented as a literal second-brain layer where conversations update local notes, principles, weekly reports, conflicts, and changes in cognition.
- AI-Assisted Reading changes reading from extracting structures by personal brainpower into using AI to make books thin, make books thick, and provoke better questions.
- AI Use Pacing is necessary because high-flow AI work can pull the person away from meals, sleep, body, and slower offline life.
- AI Organization Design needs a future “AI-era Drucker” because companies may shift from buying human brainpower to managing large numbers of agents.
- Water And Fire Education contrasts industrial “water” education that fills a container with future “fire” education that helps each child find and ignite personal will, talent, and curiosity with AI support.
- Feed Curation is treated as agency infrastructure: constraining WeChat friends, public accounts, RSS, and paper intake can increase signal density and thinking freedom.
Key Quotes
“人何以自处” — the episode’s central question for the AI era.
“脑力交给模型，心力留给人” — the compact version of Li’s human/AI boundary.
“your feed is your fate” — Meng Yan’s summary of the feed-curation thread.
Connections
- 无人知晓 — show context shared with E42, E43, and E44.
- Meng Yan / 孟岩 and Li Jigang / 李继刚 — speakers connecting investing, AI, reading, education, and selfhood.
- AI As Time Compression, AI Company Deep Well, Prompt As Intent Transmission, AMV Prompt Framework, Wet-State Human Agency, Feed Curation, and Water And Fire Education — main concept pages added by this source.
- Human Agency Under AI, Subjectivity As AI Asset, Human-Machine Amplification, AI Use Pacing, and Attention Industrialization — existing human-agency and attention threads strengthened by the episode.
- Context Engineering, Persistent Agent Memory, Personal Knowledge Ecology, AI-Assisted Reading, Reading As Frame Training, and AI Communication Ability — knowledge, prompt, and memory practices extended by Li’s workflow.
- AI Organization Design, Agentic Workflow, AI Skills, Digital Employees, and One-Person Company — organizational questions raised by the claim that companies may buy token output instead of human brainpower.
- Financial AI Agents, Investment Risk Management, Circle Of Competence, and Trust As Business Asset — finance and AI-advice implications.
- Embodied Judgment, Flow Environment Design, and Trained Intuition — adjacent pages for body, rhythm, and non-verbal judgment.
Contradictions
- Apparent tension with AI Literacy Against Worship and Human Judgment Under AI: the episode is enthusiastic about handing dry knowledge work to AI, but it does not make AI authoritative. It keeps the human responsible for intention, value choice, taste, embodied life, and final judgment.