Intelligence Devaluation

Intelligence devaluation is the Keji Luandun episode’s thesis that AI can reduce the market scarcity of education, coding ability, professional knowledge, and other cognitive skills that previously supported middle-class income and status. In 智力贬值的春节见闻录，与那场正在酝酿的优贷危机, the hosts do not argue that intelligence becomes useless; they argue that many once-premium capabilities may become cheaper to buy, automate, or approximate.

The concept is broader than AI Content Devaluation. Content devaluation concerns audience attention and generated media; intelligence devaluation concerns labor-market bargaining power, school and credential value, software work, white-collar income stability, and the credit assumptions built on that stability.

136. 全球大模型季报第9集：和广密聊，Coding是AGI第二幕、硅谷御三家真相、模型正成为新一代OS adds the coding-agent version. The source argues that when Claude Code, Codex, and similar systems perform high-value digital work, knowledge and software labor can be compressed into model capability, compute, and Token Usage. That makes the pressure more direct for junior developers, white-collar entrants, consultants, SaaS labor, and outsourcing providers.

Key Claims

• AI can turn some formerly scarce cognitive work into commodity output, reducing wage and status premiums attached to formal education or technical skill.
• Coding is the most visible example: implementation may become easier, while product judgment, deployment, verification, and business know-how remain scarce.
• “AI equality” can have two sides: it may help less-credentialed people produce more, while reducing the premium held by people whose advantage was mostly credentialed cognitive output.
• If everyone uses similar models and tools, output may become more homogeneous, making taste, field knowledge, social trust, and distribution more important.
• Human value shifts toward observing real situations, eliciting tacit needs, handling people, and deciding what should be built or said.
• The concept creates downstream household-finance risk because professional income stability was part of how many middle-class borrowers were evaluated.
• Coding-agent adoption can make intelligence devaluation arrive through real work automation rather than only through cheaper content or advice.

Connections

• Prime Borrower Credit Risk — credit-model implication when knowledge-worker income becomes less stable.
• Human Resource Deflation Compute Infrastructure Inflation — macro version where AI deflates labor costs while increasing infrastructure demand.
• AI Programming Engine Shift, Vibe Coding, and AI Engineering Thinking — software-work mechanism behind the repricing.
• Middle-Class Consumption Pressure and Financial Career Risk — household and career pressure if professional income expectations reset.
• Domain Expert Alignment, Human Judgment Under AI, and AI Communication Ability — skills that become more valuable when generic cognition is cheaper.
• AI Content Devaluation — adjacent attention and media-output devaluation.
• AGI Three Acts, AI Programming Engine Shift, Human Resource Deflation Compute Infrastructure Inflation, and Model As Operating System — coding-agent and platform-scale extension added by episode 136.