Problem Solving Through Human-AI Preference-Based Cooperation

TL;DR

This paper introduces HAICo2, a framework for human-AI cooperation in complex problem solving, addressing current AI limitations by enabling better preference expression and adaptive collaboration.

Contribution

The paper proposes a novel framework, HAICo2, for human-AI co-construction, formalizes its concepts, and discusses open research challenges.

Findings

Identifies limitations of current generative AI in complex problem solving.

Proposes a formal framework for human-AI cooperative problem solving.

Highlights open research problems in developing adaptive human-AI collaboration.

Abstract

While there is a widespread belief that artificial general intelligence (AGI) -- or even superhuman AI -- is imminent, complex problems in expert domains are far from being solved. We argue that such problems require human-AI cooperation and that the current state of the art in generative AI is unable to play the role of a reliable partner due to a multitude of shortcomings, including difficulty to keep track of a complex solution artifact (e.g., a software program), limited support for versatile human preference expression and lack of adapting to human preference in an interactive setting. To address these challenges, we propose HAICo2, a novel human-AI co-construction framework. We take first steps towards a formalization of HAICo2 and discuss the difficult open research problems that it faces.