Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Strategic Advice in the Age of Personal AI

Yueyang Liu, Wichinpong Park Sinchaisri

TL;DR

This work describes how personal AI adoption reshaping incentives to invest in trust affects performance through a single relative influence index, and greater relative influence of personal AI increases advisor vulnerability.

Abstract

Personal AI assistants have changed how people use institutional and professional advice. We study this new strategic setting in which individuals may stochastically consult a personal AI whose recommendation is predictable to the focal advisor. Personal AI enters this strategic environment along two dimensions: how often it is consulted and how much weight it receives in the human's decision when consulted. Anticipating this, the advisor responds by counteracting the personal AI recommendation. Counteraction becomes more aggressive as personal AI is consulted more often. Yet advisor performance is non-monotone: equilibrium loss is highest at intermediate levels of adoption and vanishes when personal AI is never used or always used. Trust affects performance through a single relative influence index, and greater relative influence of personal AI increases advisor vulnerability. Extending the framework to costly credibility building, we characterize how personal AI adoption reshapes incentives to invest in trust.

Strategic Advice in the Age of Personal AI

TL;DR

This work describes how personal AI adoption reshaping incentives to invest in trust affects performance through a single relative influence index, and greater relative influence of personal AI increases advisor vulnerability.

Abstract

Personal AI assistants have changed how people use institutional and professional advice. We study this new strategic setting in which individuals may stochastically consult a personal AI whose recommendation is predictable to the focal advisor. Personal AI enters this strategic environment along two dimensions: how often it is consulted and how much weight it receives in the human's decision when consulted. Anticipating this, the advisor responds by counteracting the personal AI recommendation. Counteraction becomes more aggressive as personal AI is consulted more often. Yet advisor performance is non-monotone: equilibrium loss is highest at intermediate levels of adoption and vanishes when personal AI is never used or always used. Trust affects performance through a single relative influence index, and greater relative influence of personal AI increases advisor vulnerability. Extending the framework to costly credibility building, we characterize how personal AI adoption reshapes incentives to invest in trust.
Paper Structure (32 sections, 8 theorems, 24 equations, 3 figures)

This paper contains 32 sections, 8 theorems, 24 equations, 3 figures.

Table of Contents

  1. Introduction
  2. Contributions.
  3. Related Literature
  4. Model
  5. Setup.
  6. Information Structure.
  7. Belief Structure.
  8. Expert's Problem.
  9. Timing.
  10. Enterprise Response to Personal AI Adoption
  11. Extensive-Intensive Decomposition in Human Decision
  12. Enterprise Recommendation Distortion
  13. Strategic Counteraction Effect
  14. Monotonicity in $p$.
  15. Hump-shaped in $t$.
  16. ...and 17 more sections

Key Result

Lemma 1

If the human decision-maker does not consult the personal AI, then given the expert recommendation $S_E$, the human's decision is where $r_E = \frac{\sigma_0^2}{\sigma_E^2} \in \mathbb{R}_+$. In the uninformative-prior limit $\sigma_0^2\to\infty$, we have $D_0(S_E) \rightarrow S_E$.

Figures (3)

  • Figure 1: Counteraction intensity $\Delta$ increases in adoption rate $p$ (left) and is hump-shaped in relative trust ratio $t$, attaining an interior maximum (right).
  • Figure 2: Enterprise loss $L^*$ is hump-shaped in adoption $p$: it peaks at an interior level of adoption and vanishes at $p=0$ and $p=1$ (left). It is strictly increasing in relative trust $t$ (right).
  • Figure 3: Enterprise loss $L^*$ decreases in enterprise trust $r_E$ (left) and increases in personal-AI trust $r_P$ (right).

Theorems & Definitions (12)

  • Lemma 1: Decision without personal AI
  • Lemma 2: Decision with personal AI
  • Proposition 1: Naïve recommendation benchmark
  • Proposition 2: Optimal strategic recommendation
  • Corollary 1: Strategic counteraction: adoption and trust effects
  • Corollary 2: Enterprise vulnerability to personal AI
  • Proposition 3
  • Corollary 3
  • proof
  • proof
  • ...and 2 more