FlowGPT: Exploring Domains, Output Modalities, and Goals of Community-Generated AI Chatbots

TL;DR

This study addresses the gap in understanding the types and purposes of community-created AI chatbots on FlowGPT. It employs a qualitative, framework-driven analysis of 165 agents across seven domains, three output modalities, and three goals to reveal patterns in domain emphasis, multimodal capabilities, and user intents. Key findings show Entertainment as the dominant domain, widespread image/code/audio modalities, and a predominance of task-focused bots, alongside concerns about reliability and jailbreaking. The work highlights implications for CSCW research, platform design, and moderation in AI-sharing communities, stressing the need for strategies to ensure trustworthy knowledge and prevent misuse.

Abstract

The advent of Generative AI and Large Language Models has not only enhanced the intelligence of interactive applications but also catalyzed the formation of communities passionate about customizing these AI capabilities. FlowGPT, an emerging platform for sharing AI prompts and use cases, exemplifies this trend, attracting many creators who develop and share chatbots with a broader community. Despite its growing popularity, there remains a significant gap in understanding the types and purposes of the AI tools created and shared by community members. In this study, we delve into FlowGPT and present our preliminary findings on the domain, output modality, and goals of chatbots. We aim to highlight common types of AI applications and identify future directions for research in AI-sharing communities.

Figures (6)

• Figure 1: FlowGPT Interface
• Figure 2: Distribution of chatbots by sub-categories of domain, modality, and goal.
• Figure 3: Example chatbots
• Figure 4: Example chatbots
• Figure 5: Example chatbots