Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Bounded by Risk, Not Capability: Quantifying AI Occupational Substitution Rates via a Tech-Risk Dual-Factor Model

Shuyao Gao, Minghao Huang

Abstract

The deployment of Large Language Models (LLMs) has ignited concerns about technological unemployment. Existing task-based evaluations predominantly measure theoretical "exposure" to AI capabilities, ignoring critical frictions of real-world commercial adoption: liability, compliance, and physical safety. We argue occupations are not eradicated instantaneously, but gradually encroached upon via atomic actions. We introduce a Tech-Risk Dual-Factor Model to re-evaluate this. By deconstructing 923 occupations into 2,087 Detailed Work Activities (DWAs), we utilize a multi-agent LLM ensemble to score both technical feasibility and business risk. Through variance-based Human-in-the-Loop (HITL) validation with an expert panel, we demonstrate a profound cognitive gap: isolated algorithmic probabilities fail to encapsulate the "institutional premium" imposed by experts bounded by professional liability. Applying a strictly algorithmic baseline via mathematical bottleneck aggregation, we calculate Relative Occupational Automation Indices ($OAI$) for the U.S. labor market. Our findings challenge the traditional Routine-Biased Technological Change (RBTC) hypothesis. Non-routine cognitive roles highly dependent on symbolic manipulation (e.g., Data Scientists) face unprecedented exposure ($OAI \approx 0.70$). Conversely, unstructured physical trades and high-stakes caretaking roles exhibit absolute resilience, quantifying a profound "Cognitive Risk Asymmetry." We hypothesize the emergent necessity of a "Compliance Premium," indicating wage resilience increasingly tied to risk-absorption capacity. We frame these findings as a cross-sectional diagnostic of systemic vulnerability, establishing a foundation for subsequent Computable General Equilibrium (CGE) econometric modeling involving dynamic wage elasticity and structural labor reallocation.

Bounded by Risk, Not Capability: Quantifying AI Occupational Substitution Rates via a Tech-Risk Dual-Factor Model

Abstract

The deployment of Large Language Models (LLMs) has ignited concerns about technological unemployment. Existing task-based evaluations predominantly measure theoretical "exposure" to AI capabilities, ignoring critical frictions of real-world commercial adoption: liability, compliance, and physical safety. We argue occupations are not eradicated instantaneously, but gradually encroached upon via atomic actions. We introduce a Tech-Risk Dual-Factor Model to re-evaluate this. By deconstructing 923 occupations into 2,087 Detailed Work Activities (DWAs), we utilize a multi-agent LLM ensemble to score both technical feasibility and business risk. Through variance-based Human-in-the-Loop (HITL) validation with an expert panel, we demonstrate a profound cognitive gap: isolated algorithmic probabilities fail to encapsulate the "institutional premium" imposed by experts bounded by professional liability. Applying a strictly algorithmic baseline via mathematical bottleneck aggregation, we calculate Relative Occupational Automation Indices () for the U.S. labor market. Our findings challenge the traditional Routine-Biased Technological Change (RBTC) hypothesis. Non-routine cognitive roles highly dependent on symbolic manipulation (e.g., Data Scientists) face unprecedented exposure (). Conversely, unstructured physical trades and high-stakes caretaking roles exhibit absolute resilience, quantifying a profound "Cognitive Risk Asymmetry." We hypothesize the emergent necessity of a "Compliance Premium," indicating wage resilience increasingly tied to risk-absorption capacity. We frame these findings as a cross-sectional diagnostic of systemic vulnerability, establishing a foundation for subsequent Computable General Equilibrium (CGE) econometric modeling involving dynamic wage elasticity and structural labor reallocation.

Paper Structure

This paper contains 25 sections, 4 equations, 4 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Literature Review
  3. The Task-Based Approach and Routine-Biased Technological Change (RBTC)
  4. The LLM Shock: Targeting Non-Routine Cognitive Work
  5. Risk Aversion, Hallucinations, and Moravec's Paradox
  6. Methodology and Data Pipeline
  7. Data Foundation: Deconstructing Occupations into Atomic Actions
  8. AI-Driven Mass Scoring: The Multi-Model Ensemble
  9. Human-in-the-Loop Validation: A Stratified 3$\times$3 Matrix Analysis
  10. Finalization of DWA Scores: Non-Linear Risk and the "No Double Penalization" Rationale
  11. The Occupational Replaceability Model
  12. Construction of the Automation Index via Tech-Risk Dual-Factor Mapping
  13. Upward Aggregation: From Detailed Work Activities to Occupational Replaceability
  14. Industry-wide Replaceability Ranking and Trait Analysis
  15. Robustness Check: Sensitivity of the Dual-Factor Matrix
  16. ...and 10 more sections

Figures (4)

  • Figure 1: Distribution of 2,087 Detailed Work Activities (DWAs) across the orthogonal dimensions of Technical Capability and Business Risk. The bubble area represents the frequency of atomic actions, with a specific color gradient reflecting the escalation of risk. The dense clustering in medium-to-high risk zones ($R \ge 3$) underscores the friction of real-world AI deployment.
  • Figure 2: Visualizing the Cognitive Gap in Risk Perception. In the Consensus Zone, human experts and AI models align closely. However, in the Severe Divergence Zone, human management experts exhibit strong loss aversion, inflating the perceived risk score by +0.35 compared to the objective AI baseline. This asymmetry highlights the friction between purely statistical capabilities and rational institutional risk pricing.
  • Figure 3: The Tech-Risk Dual-Factor Automation Matrix. The color gradient represents the Automation Index ($AI$), illustrating the non-linear penalization of technical capabilities by business, legal, and safety risks.
  • Figure 4: Macroeconomic Labor Market Vulnerability Distribution of the Occupational Automation Index (OAI). The density plot highlights that full replacement is an illusion for the majority of occupations, with exposure concentrated in specific cognitive domains rather than a uniform market-wide displacement.