Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Social Skill Training with Large Language Models

Diyi Yang, Caleb Ziems, William Held, Omar Shaikh, Michael S. Bernstein, John Mitchell

TL;DR

Social skills are critical for workplace and personal success but scalable training opportunities are scarce. The authors propose APAM, a framework that pairs an AI Partner for realistic practice with an AI Mentor for tailored, knowledge-grounded feedback, leveraging LLM-based simulations and retrieval-augmented knowledge. They instantiate the framework with CARE, Rehearsal, and GPTeach to demonstrate practical applications across counseling, conflict resolution, and education, and outline a safety continuum for deployment. The work advocates cross-disciplinary collaboration to harness AI for workforce development and greater social equity while acknowledging safety and bias challenges.

Abstract

People rely on social skills like conflict resolution to communicate effectively and to thrive in both work and personal life. However, practice environments for social skills are typically out of reach for most people. How can we make social skill training more available, accessible, and inviting? Drawing upon interdisciplinary research from communication and psychology, this perspective paper identifies social skill barriers to enter specialized fields. Then we present a solution that leverages large language models for social skill training via a generic framework. Our AI Partner, AI Mentor framework merges experiential learning with realistic practice and tailored feedback. This work ultimately calls for cross-disciplinary innovation to address the broader implications for workforce development and social equality.

Social Skill Training with Large Language Models

TL;DR

Social skills are critical for workplace and personal success but scalable training opportunities are scarce. The authors propose APAM, a framework that pairs an AI Partner for realistic practice with an AI Mentor for tailored, knowledge-grounded feedback, leveraging LLM-based simulations and retrieval-augmented knowledge. They instantiate the framework with CARE, Rehearsal, and GPTeach to demonstrate practical applications across counseling, conflict resolution, and education, and outline a safety continuum for deployment. The work advocates cross-disciplinary collaboration to harness AI for workforce development and greater social equity while acknowledging safety and bias challenges.

Abstract

People rely on social skills like conflict resolution to communicate effectively and to thrive in both work and personal life. However, practice environments for social skills are typically out of reach for most people. How can we make social skill training more available, accessible, and inviting? Drawing upon interdisciplinary research from communication and psychology, this perspective paper identifies social skill barriers to enter specialized fields. Then we present a solution that leverages large language models for social skill training via a generic framework. Our AI Partner, AI Mentor framework merges experiential learning with realistic practice and tailored feedback. This work ultimately calls for cross-disciplinary innovation to address the broader implications for workforce development and social equality.
Paper Structure (34 sections, 1 figure, 2 tables)

This paper contains 34 sections, 1 figure, 2 tables.

Table of Contents

  1. Introduction
  2. LLMs for Characters and Simulation
  3. The APAM Framework
  4. AI Partner
  5. AI Mentor
  6. Methodology
  7. Examples of APAM
  8. CARE (AI Mentor)
  9. Rehearsal (AI Partner)
  10. GPTeach (AI Partner)
  11. Vision for Safe Deployment
  12. AI Partner Continuum
  13. Mode 1: Rubber Ducking
  14. Mode 2: Peer Roleplay
  15. Mode 3: Standardized Partner
  16. ...and 19 more sections

Figures (1)

  • Figure 1: Modes of the APAM framework. As AI capabilities improve, the APAM framework develops from its basis in non-AI teaching practices towards the possibility of realistic simulated AI Partner learning scenarios augmented with AI Mentor feedback that can be personalized based on prior practice sessions between the User and the AI Partner. With LLMs, our prior work has shown that AI Mentors can effectively generate suggestions based on best practices hsu2023helping and AI Partners can replicate many of the benefits of roleplay shaikh2023rehearsal.