Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Human vs. Generative AI in Content Creation Competition: Symbiosis or Conflict?

Fan Yao, Chuanhao Li, Denis Nekipelov, Hongning Wang, Haifeng Xu

TL;DR

The paper introduces a generalized Tullock contest to study competition between human creators and GenAI across multiple topics, modeling GenAI as an external influencer in exclusive settings and as an optional tool in inclusive settings. It derives rigorous equilibrium results: a unique pure Nash equilibrium exists for the exclusive competition under mild conditions, while the inclusive competition may lack a PNE in general, though a tractable 1-D separable-cost case does admit structured equilibria with GenAI adoption. The authors provide macro-scale insights into content volume via s^* and show that increased GenAI power or market size can drive more creators toward GenAI, yet a symbiotic equilibrium where human creators persist remains feasible. They validate their theory with experiments using PNE solvers, illustrating how GenAI capability reshapes content allocation and individual utilities, with implications for platform design and policy to foster fair and diverse content ecosystems.

Abstract

The advent of generative AI (GenAI) technology produces transformative impact on the content creation landscape, offering alternative approaches to produce diverse, high-quality content across media, thereby reshaping online ecosystems but also raising concerns about market over-saturation and the potential marginalization of human creativity. Our work introduces a competition model generalized from the Tullock contest to analyze the tension between human creators and GenAI. Our theory and simulations suggest that despite challenges, a stable equilibrium between human and AI-generated content is possible. Our work contributes to understanding the competitive dynamics in the content creation industry, offering insights into the future interplay between human creativity and technological advancements in GenAI.

Human vs. Generative AI in Content Creation Competition: Symbiosis or Conflict?

TL;DR

The paper introduces a generalized Tullock contest to study competition between human creators and GenAI across multiple topics, modeling GenAI as an external influencer in exclusive settings and as an optional tool in inclusive settings. It derives rigorous equilibrium results: a unique pure Nash equilibrium exists for the exclusive competition under mild conditions, while the inclusive competition may lack a PNE in general, though a tractable 1-D separable-cost case does admit structured equilibria with GenAI adoption. The authors provide macro-scale insights into content volume via s^* and show that increased GenAI power or market size can drive more creators toward GenAI, yet a symbiotic equilibrium where human creators persist remains feasible. They validate their theory with experiments using PNE solvers, illustrating how GenAI capability reshapes content allocation and individual utilities, with implications for platform design and policy to foster fair and diverse content ecosystems.

Abstract

The advent of generative AI (GenAI) technology produces transformative impact on the content creation landscape, offering alternative approaches to produce diverse, high-quality content across media, thereby reshaping online ecosystems but also raising concerns about market over-saturation and the potential marginalization of human creativity. Our work introduces a competition model generalized from the Tullock contest to analyze the tension between human creators and GenAI. Our theory and simulations suggest that despite challenges, a stable equilibrium between human and AI-generated content is possible. Our work contributes to understanding the competitive dynamics in the content creation industry, offering insights into the future interplay between human creativity and technological advancements in GenAI.
Paper Structure (20 sections, 12 theorems, 69 equations, 8 figures, 4 algorithms)

This paper contains 20 sections, 12 theorems, 69 equations, 8 figures, 4 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Modeling Content Creation Competition between Humans and GenAI
  4. Context: Exclusive and Inclusive Competitions
  5. The Case of Separable Costs and $1$-D Competition
  6. Exclusive Human-vs-GenAI Competitions
  7. Equilibrium Properties of the 1-D Competition
  8. Inclusive Human-vs-GenAI Competitions
  9. Equilibrium Properties of the 1-D competition
  10. Experiments
  11. Conclusion
  12. Proof of Theorem \ref{['thm:monotone_G']}
  13. Proof of Theorem \ref{['lm:order_c']}
  14. Proof of Proposition \ref{['lm:cost_bound']}
  15. Proof of Theorem \ref{['thm:asym_s']}
  16. ...and 5 more sections

Key Result

Theorem 1

Consider any $\mathcal{G}_{EX}$$(\bm{\alpha}, \bm{\beta},\bm{\gamma},\bm{\mu},\{c_i\}_{i=1}^n)$. If $\bm{\beta} \in [0, 1]^K$, then the game is a strictly monotone game hence admits a unique pure Nash Equilibrium.

Figures (8)

  • Figure 1: The total body of content creation $s^*$ at the PNE as a function of $\mu,n$ (Left) and $\alpha,\beta$ (Right).
  • Figure 2: The total human utilities $W=\sum_i u_i$ at the PNE as a function of $\mu,n$ (Left) and $\alpha,\beta$ (Right). Default $(\alpha,\beta)=(1.0, 0.5)$.
  • Figure 4: The percentage of GenAI creators at PNE of $\mathcal{G}^{(1)}_{IN}$ under different $\alpha,\beta$ or $\mu,n$.
  • Figure 5: How likely a creator with cost ranked at different percentiles tend to switch to GenAI at an arbitrary PNE of $\mathcal{G}^{(1)}_{IN}$.
  • Figure 6: Left/Right: Total/average utility and content creation of human creators vs GenAI creators.
  • ...and 3 more figures

Theorems & Definitions (15)

  • Remark 1
  • Definition 1
  • Remark 2
  • Theorem 1
  • Theorem 2: Micro-level Equilibrium Properties
  • Proposition 1: The utility--cost balance at equilibrium
  • Theorem 3: Macro-level Equilibrium Properties
  • Theorem 4
  • Theorem 5
  • Lemma 1
  • ...and 5 more