Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

From Augmentation to Symbiosis: A Review of Human-AI Collaboration Frameworks, Performance, and Perils

Richard Jiarui Tong

TL;DR

The paper surveys a 60-year trajectory of human-AI collaboration, tracing the historical tension between treating AI as a cooperative teammate (Licklider) versus a powerful augmentation tool (Engelbart) and advocating a modern shift toward human-centered, tool-like AI (Shneiderman). It formalizes a causal chain—Explainable AI (XAI) enabling Shared Mental Models (SMMs) through a co-adaptive loop—to explain the observed performance paradox: negative synergy in decision tasks but positive synergy in content creation, with Centaur illustrating model-guided discovery that aligns with formulative goals. It identifies psychological and ethical barriers, notably the Algorithm-in-the-Loop problem, cognitive biases (algorithm aversion vs automation bias), and cognitive deskilling, calling for transparent governance and robust XAI. The work proposes a unifying theoretical frame—dual-process cognition plus the extended self—to internalize AI into human cognition and achieve a unitary human-XAI symbiotic agency, outlining longitudinal research directions and design principles emphasizing formulation over sole decision-making.

Abstract

This paper offers a concise, 60-year synthesis of human-AI collaboration, from Licklider's ``man-computer symbiosis" (AI as colleague) and Engelbart's ``augmenting human intellect" (AI as tool) to contemporary poles: Human-Centered AI's ``supertool" and Symbiotic Intelligence's mutual-adaptation model. We formalize the mechanism for effective teaming as a causal chain: Explainable AI (XAI) -> co-adaptation -> shared mental models (SMMs). A meta-analytic ``performance paradox" is then examined: human-AI teams tend to show negative synergy in judgment/decision tasks (underperforming AI alone) but positive synergy in content creation and problem formulation. We trace failures to the algorithm-in-the-loop dynamic, aversion/bias asymmetries, and cumulative cognitive deskilling. We conclude with a unifying framework--combining extended-self and dual-process theories--arguing that durable gains arise when AI functions as an internalized cognitive component, yielding a unitary human-XAI symbiotic agency. This resolves the paradox and delineates a forward agenda for research and practice.

From Augmentation to Symbiosis: A Review of Human-AI Collaboration Frameworks, Performance, and Perils

TL;DR

The paper surveys a 60-year trajectory of human-AI collaboration, tracing the historical tension between treating AI as a cooperative teammate (Licklider) versus a powerful augmentation tool (Engelbart) and advocating a modern shift toward human-centered, tool-like AI (Shneiderman). It formalizes a causal chain—Explainable AI (XAI) enabling Shared Mental Models (SMMs) through a co-adaptive loop—to explain the observed performance paradox: negative synergy in decision tasks but positive synergy in content creation, with Centaur illustrating model-guided discovery that aligns with formulative goals. It identifies psychological and ethical barriers, notably the Algorithm-in-the-Loop problem, cognitive biases (algorithm aversion vs automation bias), and cognitive deskilling, calling for transparent governance and robust XAI. The work proposes a unifying theoretical frame—dual-process cognition plus the extended self—to internalize AI into human cognition and achieve a unitary human-XAI symbiotic agency, outlining longitudinal research directions and design principles emphasizing formulation over sole decision-making.

Abstract

This paper offers a concise, 60-year synthesis of human-AI collaboration, from Licklider's ``man-computer symbiosis" (AI as colleague) and Engelbart's ``augmenting human intellect" (AI as tool) to contemporary poles: Human-Centered AI's ``supertool" and Symbiotic Intelligence's mutual-adaptation model. We formalize the mechanism for effective teaming as a causal chain: Explainable AI (XAI) -> co-adaptation -> shared mental models (SMMs). A meta-analytic ``performance paradox" is then examined: human-AI teams tend to show negative synergy in judgment/decision tasks (underperforming AI alone) but positive synergy in content creation and problem formulation. We trace failures to the algorithm-in-the-loop dynamic, aversion/bias asymmetries, and cumulative cognitive deskilling. We conclude with a unifying framework--combining extended-self and dual-process theories--arguing that durable gains arise when AI functions as an internalized cognitive component, yielding a unitary human-XAI symbiotic agency. This resolves the paradox and delineates a forward agenda for research and practice.
Paper Structure (40 sections, 2 tables)

This paper contains 40 sections, 2 tables.

Table of Contents

  1. The Symbiotic Vision: Foundational Concepts in Human-Machine Partnership
  2. Introduction: The Licklider Vision of "Man-Computer Symbiosis"
  3. The Engelbart Framework: "Augmenting Human Intellect"
  4. Synthesis: The Foundational Tension (Tool vs. Teammate)
  5. Modern Frameworks for Human-AI Interaction
  6. The Modern Lexicon: Augmented, Symbiotic, and Collective Intelligence
  7. Shneiderman's Human-Centered AI (HCAI) Framework
  8. The Two-Dimensional HCAI Framework
  9. The "Supertool" Metaphor
  10. Mechanisms for Effective Human-AI Symbiosis
  11. Introduction: The Prerequisites for Partnership
  12. Pillar 1: Explainable AI (XAI) as the Mechanism for Trust
  13. Pillar 2: Shared Mental Models (SMMs) as the Psychological Goal
  14. The XAI-SMM Causal Chain
  15. Pillar 3: Co-Adaptation as the Process for Building SMMs
  16. ...and 25 more sections