Reframing individual roles in collaboration: digital identity construction and adaptive mechanisms for resistance-based professional skills in AI-human intelligence symbiosis

Weizheng Jiang, Yongzhou Li, Xiaoling Hu, Dongling MaFrontiers in Psychology 2025 · DOI

TL;DR

This paper proposes a new framework for understanding how individuals can thrive in AI-human collaboration by focusing on "digital identity construction." It identifies "AI-resistant" skills like foreign language proficiency and passion/optimism as crucial, and suggests a personalized approach to talent development based on an individual's unique profile within this framework.

What Problem Does This Paper Address?

The rapid advancement of AI, particularly generative AI, is transforming industries. While AI excels at certain tasks, human-AI collaboration still faces challenges in achieving optimal effectiveness and innovation. Existing research often isolates professional skills from their broader context, neglecting how an individual's digital identity – shaped by their motivation, position in networks, and how they articulate knowledge – influences their performance. This oversight hinders sustainable skill development and adaptability in AI-human teams. The paper aims to bridge this gap by exploring how individual digital identity construction impacts AI-human collaborative innovation.

Methods

The study employs a multi-faceted quantitative approach:

  • Triadic Synergy Framework: This framework conceptualizes digital identity as a product of three interacting dimensions: intrinsic motivation (proactive personality - PP), structural positioning (network centrality - NC), and knowledge articulation (professional skills).
  • Entropy Weight Method: Used to objectively assign weights to different cognitive levels (based on Bloom's taxonomy) in knowledge conversion, reflecting their contribution to AI-HI collaboration effectiveness.
  • Gradient Descent (GD) Algorithm: Applied to model complex, non-linear relationships between professional skills, network centrality, proactive personality, and collaborative innovation effectiveness. GD is chosen for its ability to handle high-dimensional data and avoid methodological disputes associated with traditional regression.
  • Residual-Matching Analysis: A novel approach to quantify the discrepancy between predicted and actual digital identity expression. This is used to classify individuals into distinct groups (core innovators, marginal experts, low performers) and inform personalized adaptation strategies.

Datasets

The study utilized data from 418 participants in the financial co-production sector in Wuhan, China, collected between 2022 and 2024. This dataset included information on proactive personality traits, community network structures, and cognitive skills. No new benchmarks were introduced; rather, existing scales and methods were adapted and integrated.

Key Insights and Ideas

  • Beyond Hard vs. Soft Skills: The paper argues for a more nuanced view of "AI-resistant" skills, moving beyond the traditional hard/soft dichotomy. It highlights that certain soft skills (like passion/optimism) and a specific hard skill (foreign language proficiency) are more resilient to AI replacement.
  • Digital Identity is Key: An individual's "digital identity" – how they are recognized and position themselves within AI-human systems – is crucial for effective collaboration. This identity is not solely system-assigned but co-constructed.
  • Synergy Matters: The interplay of intrinsic motivation (proactive personality), structural position (network centrality), and knowledge articulation (skills) significantly shapes digital identity and, consequently, collaborative innovation.
  • Empowerment vs. Marginalization: Network centrality can either empower or marginalize an individual's digital identity and skill expression, depending on skill fit and personal motivation.
  • Personalized Talent Development: A one-size-fits-all approach to talent development is insufficient. Individuals can be categorized into distinct profiles (core innovators, marginal experts, low performers) requiring tailored strategies.

Key Results

  • AI-Resistant Skills: Foreign Language Proficiency (FL) and Passion/Optimism (PO) were identified as the most "AI-resistant" professional skills, showing significant positive impacts on AI-HI collaboration effectiveness. PO had a coefficient of 0.03 and FL 0.07 in initial regressions, and remained significant (PO: 0.04, FL: 0.03) even when controlling for network centrality and proactive personality.
  • Impact of NC and PP: Network Centrality (NC; β = 0.04) and Proactive Personality (PP; β = 0.05) significantly amplify the impact of these AI-resistant skills on collaboration effectiveness.
  • Cognitive Contributions: Higher-order cognitive processes (creation, evaluation, analysis) play a more significant role in AI-HI collaborative innovation than lower-order ones, reflecting AI's substitution effect.
  • Digital Identity Classification: The study successfully classified participants into three groups: Core Innovators (high match, low residual), Marginal Experts (medium match, high residual), and Low-efficiency Individuals (low match, low residual), with distinct adaptation recommendations for each.

Limitations and Future Work

  • Subjectivity of Soft Skills: The paper acknowledges that soft skill assessment, relying on questionnaires, remains inherently subjective. Future research should explore more objective methods, like technology-driven assessments.
  • Model Refinement: While the triadic synergy model is effective, variables like AI model performance and specific task scenarios could be further refined.
  • Qualitative Data Integration: The study primarily used quantitative methods. Incorporating qualitative data (e.g., interviews) could enhance explanatory power.
  • Cross-Sectional Design: The current study is cross-sectional, limiting the understanding of AI technology evolution and its impact over time. Longitudinal studies are recommended.
  • AI System Characteristics: The study did not quantitatively assess the technical characteristics of AI systems, which are known to influence digital identity construction and task adaptation. Future work should aim to quantify these features.
  • Sample Homogeneity: While substantial, the sample was drawn from a specific financial co-production sector and may not be fully representative. Cross-validation across diverse industries and formal employees is suggested.

Summary updated

Abstract

Amid the unprecedented wave of AI advancement, AI-resistant professional skills play a significant role in enhancing the effectiveness of human-AI collaboration. However, existing research tends to isolate professional skills from their broader context, overlooking the triadic construction of digital identity recognition through individual motivation, structural position, and knowledge articulation. This oversight weakens the sustainability and adaptability of skill expression, thereby hindering innovation performance in AI-HI (Artificial Intelligence-Human Intelligence) collaboration. Drawing on the entropy weight method, gradient descent algorithm, and a residual-matching decision matrix, this study conducted quantitative modeling of 418 participants in the financial co-production sector from 2022 to 2024. The findings reveal that network centrality (NC; β = 0.04<sup>**</sup>) and proactive personality (PP; β = 0.05<sup>**</sup>) significantly amplify the impact of two key AI-resistant skills-foreign language proficiency (FL) and passion/optimism (PO)-on collaboration effectiveness, through structural empowerment and intrinsic motivation. Furthermore, this study develops a digital identity recognition and classification framework that identifies three distinct groups: core innovators, marginal experts, and low performers. By extending the theoretical model of digital identity construction within AI-HI collaboration, this study also proposes a differentiated approach to talent development and resource allocation based on innovation effectiveness and identity alignment, offering new insights into the advancement of digital human capital.
BibTeX
@inproceedings{jiang2025reframing,
  author = {Weizheng Jiang and Yongzhou Li and Xiaoling Hu and Dongling Ma},
  title = {{Reframing individual roles in collaboration: digital identity construction and adaptive mechanisms for resistance-based professional skills in AI-human intelligence symbiosis}},
  booktitle = {Frontiers in Psychology},
  year = {2025},
  doi = {10.3389/fpsyg.2025.1652130},
}
Figure