Simple autonomous agents can enhance creative semantic discovery by human groups

Area of Science:

• Cognitive Science
• Social Psychology
• Artificial Intelligence

Background:

• Groups can foster innovation better than individuals, but face challenges like groupthink.
• Understanding how to optimize group innovation is crucial for scientific and technological advancement.

Purpose of the Study:

• To investigate how networked groups and autonomous agents (bots) influence the exploration of semantic spaces for innovation.
• To determine the impact of different bot strategies on group creative discovery.

Main Methods:

• Experiments involved 1875 participants in 125 networks of 15 nodes, searching a semantic space of 20,000 nouns.
• Bots employed strategies of sharing random, most similar, or least similar nouns from neighbors.
• Performance was compared between individuals, social groups, and groups with bots.

Main Results:

• Groups demonstrated superior exploration of the semantic space compared to isolated individuals.
• Groups with bots sharing the most similar nouns showed enhanced performance in easily navigable semantic spaces.
• Simple, interpretable bot behaviors can significantly impact human group creative discovery.

Conclusions:

• Networked human groups are effective at exploring complex idea spaces.
• Strategic integration of autonomous agents can augment human group innovation capacity.
• This research offers insights into human-AI collaboration for enhanced creativity.