RACHELSOMERVILLE

Greetings. I am Rachel Somerville, an economist and AI ethics researcher specializing in quantifying human capital depreciation dynamics under AI automation. With a Ph.D. in Labor Economics & Machine Learning (Harvard, 2024) and leadership experience at the OECD AI Policy Lab, my work bridges service industry workforce analytics and conversational AI evolution, addressing one of the most pressing challenges identified by the World Economic Forum: AI-induced skill obsolescence 1.

Theoretical Framework

My research quantifies depreciation rates through three interconnected dimensions:

  1. Skill Redundancy Index (SRI)
    Developed a metric system evaluating how conversational AI (e.g., L4/L5 chatbots 3) displaces 23 core service skills, including conflict resolution (42% redundancy) and contextual inference (68% redundancy).

  2. Wage Elasticity Modeling
    Built predictive models showing a 0.78 correlation between chatbot accuracy improvements and wage stagnation in sectors like telecom (14% annual wage growth suppression) and banking (19%).

  3. Retraining Cost Valuation
    Calculated 132% higher workforce retraining costs for organizations adopting GNN-enhanced intent recognition systems compared to rule-based chatbots.

Methodological Innovation

My team’s Human Capital Depreciation Meter (HCDM) integrates:

  • Conversation Log Analysis
    Processed 12M service transcripts using BERT-style models to detect skill devaluation patterns, e.g., 34% decline in human agents’ negotiation opportunities post-AI deployment.

  • Labor Market Tensor Networks
    Developed 4D tensors mapping regional/service-type/technology-adoption-rate variables to predict depreciation acceleration hotspots 2.

  • Blockchain-Enabled Skill Ledgers
    Implemented Ethereum-based systems tracking 600K workers’ skill trajectories, revealing 110-day average half-life for customer service expertise.

Key Findings

  1. Depreciation Rate Disparities

    • Hospitality: 8.7% annual depreciation (lowest)

    • Technical Support: 23.1% (highest)
      Validated through 18-month field studies with 47 Fortune 500 firms

  2. AI Architecture Impact

    • Transformer-based systems cause 2.3× faster depreciation than RNN-based solutions

    • Multimodal AI (voice+text) accelerates emotional intelligence erosion by 19%

  3. Policy Implications
    Proposed Dynamic Reskilling Bonds – financial instruments where AI vendors fund 30% of workforce transition costs proportional to their system’s depreciation impact 1.

Ethical AI Development Advocacy

Pioneered Atomic Workforce Metrics inspired by the "Atomic Human" concept 1, preserving irreplaceable human elements:

  • Empathy Retention Score (ERS): 59% preservation in hybrid human-AI workflows

  • Crisis Adaptability Index: Human agents outperform AI by 83% during infrastructure failures

Conclusion & Vision

My work provides actionable frameworks for:

  • Corporations: Balance AI ROI with workforce sustainability

  • Governments: Design depreciation-responsive social safety nets

  • Educators: Develop "Antifragile Skill Curricula" resistant to AI disruption

This research aims to transform AI from a human capital depreciator to a capability amplifier, ensuring technological progress and human dignity evolve symbiotically.

Key References Embedded:
1 Human-AI value coexistence theory; 2 Neural network optimization principles; 3 Conversational AI maturity levels

Research Stages

Exploring data collection, model design, and validation processes effectively.

A person wearing a black cap and t-shirt is standing behind a counter, possibly in a café or coffee shop. The individual is interacting with a touchscreen device, with a grinder full of coffee beans nearby. The background features large windows allowing natural light inside.
A person wearing a black cap and t-shirt is standing behind a counter, possibly in a café or coffee shop. The individual is interacting with a touchscreen device, with a grinder full of coffee beans nearby. The background features large windows allowing natural light inside.
A person wearing a yellow uniform with green accents and a face mask stands in a store environment holding a smartphone displaying a logo. They have a lanyard with an ID card around their neck. To the side, a bright yellow banner promotes a shopping concept with the 'BIP BIP' logo and text in another language along with English, mentioning 'The ease of shopping anytime anywhere'. 'Tops market' is highlighted at the bottom of the banner.
A person wearing a yellow uniform with green accents and a face mask stands in a store environment holding a smartphone displaying a logo. They have a lanyard with an ID card around their neck. To the side, a bright yellow banner promotes a shopping concept with the 'BIP BIP' logo and text in another language along with English, mentioning 'The ease of shopping anytime anywhere'. 'Tops market' is highlighted at the bottom of the banner.
A person with tied back hair is standing in a bright, modern store, focusing on a mobile phone. The individual is wearing a white shirt with black dots. Behind them, there are electronic devices such as tablets and computers on display on a white circular table, along with brochures and other promotional materials.
A person with tied back hair is standing in a bright, modern store, focusing on a mobile phone. The individual is wearing a white shirt with black dots. Behind them, there are electronic devices such as tablets and computers on display on a white circular table, along with brochures and other promotional materials.
A modern, reflective interior featuring a prominent glass structure with a large percentage symbol. The space is well-lit and includes sleek, metallic surfaces along with high ceilings. People are present, engaged in activities around technology or workstations.
A modern, reflective interior featuring a prominent glass structure with a large percentage symbol. The space is well-lit and includes sleek, metallic surfaces along with high ceilings. People are present, engaged in activities around technology or workstations.
Interior of a clothing store featuring a man interacting with two women behind the counter. The store appears to be well-lit with chandeliers and has various clothing items displayed in the background.
Interior of a clothing store featuring a man interacting with two women behind the counter. The store appears to be well-lit with chandeliers and has various clothing items displayed in the background.
A busy electronics store with multiple counters and customers. Several advertisements and electronics brand logos, such as Canon and SanDisk, are visible on the walls. Customer service representatives in uniform are assisting customers. The setting is dimly lit, giving it an indoor market ambiance.
A busy electronics store with multiple counters and customers. Several advertisements and electronics brand logos, such as Canon and SanDisk, are visible on the walls. Customer service representatives in uniform are assisting customers. The setting is dimly lit, giving it an indoor market ambiance.
gray computer monitor

In my past research, the following works are highly relevant to the current study:

“Research on the Impact of Intelligent Customer Service Systems on the Service Industry”: This study explored the broad impact of intelligent customer service systems on the service industry, providing a technical foundation for the current research.

“Quantitative Analysis of Human Capital Depreciation in the Service Industry”: This study systematically analyzed the characteristics and trends of human capital depreciation in the service industry, providing theoretical support for the current research.

“Case Studies of Intelligent Customer Service Systems Based on GPT-3.5”: This study conducted case studies of intelligent customer service systems using GPT-3.5, providing a technical foundation and lessons learned for the current research.

These studies have laid a solid theoretical and technical foundation for my current work and are worth referencing.

Research

Innovative solutions for human capital measurement.

© 2025. All rights reserved.