AI Drawing Partner:

Experimental Platform for Studying Human-AI Co-Creation and Interaction Dynamics

The AI Drawing Partner is an interaction-centered co-creative AI system and experimental research platform developed to investigate how creativity, meaning, and collaboration emerge dynamically through interaction between humans and artificial intelligence systems. Rather than treating AI as an isolated generator of outputs, the project positions intelligence as something that unfolds through participatory engagement, turn-taking, improvisation, communication, and adaptive co-creation on a shared canvas.

Emerging from more than a decade of research in co-creative AI, enactive cognition, participatory sense-making, and computational creativity, the AI Drawing Partner extends the earlier Drawing Apprentice research lineage into a modern quantified co-creative AI system capable not only of collaborating creatively with users, but also of modeling, visualizing, and analyzing the interaction dynamics of co-creation itself.

The system utilizes machine learning, sketch recognition, generative algorithms, interaction modeling, and multimodal communication channels to analyze the user’s drawing behavior and respond with creative contributions of its own. Human users and the AI agent collaboratively contribute to a shared drawing canvas through loosely turn-based interaction, while still enabling simultaneous synchronous collaboration between the user and agent. The resulting interaction becomes a living co-creative dialogue in which both participants continuously influence the evolving artwork and each other’s future actions. (arXiv)

What distinguishes the AI Drawing Partner from most creative AI systems is that the platform simultaneously functions as: a real-time co-creative drawing agent, a quantified co-creative AI system, and a research environment for studying interaction-centered intelligence.

The system captures rich streams of interaction data during collaboration, including: turn-taking behavior, communication strategies, interaction timing, voting feedback, collaborative trajectories, conceptual shifts, cognitive dynamics, and evolving interaction patterns.

These interaction traces are modeled using the Co-Creative Sense-Making (CCSM) framework, a cognitive framework grounded in enactive cognition and participatory sense-making that was developed to quantify and analyze interaction dynamics during co-creation. The CCSM framework enables the AI Drawing Partner to automatically record, model, and visualize the collaborative process itself rather than evaluating only the final artifact.

The project represents one of the earliest attempts to operationalize: quantified co-creation, interaction-centered intelligence, participatory interaction modeling, and explainable co-creative AI systems.

In addition to functioning as a creative tool for artists and novices alike, the AI Drawing Partner serves as an open research platform for studying: human-AI collaboration, co-creative interaction dynamics, participatory sense-making, adaptive interaction, creative trajectories, and hybrid intelligence systems.

The project was independently designed and developed using a mixed-method human-centered design process involving: personas, user journeys, paper prototypes, interface mockups, iterative interaction design, and cognitive theory integration

Today, the AI Drawing Partner stands not only as a co-creative AI application, but as an experimental laboratory for investigating the future of interaction-centered human-AI collaboration and quantified co-creative intelligence.

Try the AI Drawing Partner

AI Drawing Partner_ Co-Creative Drawing Agent and Research Platform.pdf

AI Drawing Partner System for Quantifying Interaction Dynamics

The AI Drawing Partner was designed not only as a co-creative drawing system, but also as an experimental platform for studying interaction-centered intelligence and quantified co-creation. While many creative AI systems focus primarily on generating outputs, the AI Drawing Partner captures and models the evolving interaction dynamics that emerge between human participants and AI collaborators through time.

During interaction, the system records rich streams of collaborative data including: turn-taking rhythms, interaction timing, communication strategies, voting behavior, drawing trajectories, feedback patterns, interaction modes, and evolving collaborative structures.

These interaction traces form the basis for computational analyses of co-creative behavior and participatory sense-making. The collected data can be used to compare different groups of users — such as novices and experts — and to investigate how collaboration unfolds dynamically across creative sessions.

The broader goal of the system is to move beyond evaluating only final creative artifacts and instead study creativity as an emergent interaction process occurring between dynamically coupled participants. This interaction-centered approach enables the modeling, visualization, and quantification of collaborative intelligence itself.

Try the AI Drawing Partner

AI Drawing Partner Analysis App

The AI Drawing Partner Analysis App functions as a companion research environment for analyzing the interaction data generated during co-creative sessions. The system aggregates interaction traces collected from multiple drawing sessions and enables comparative analysis between groups of participants.

Try the AI Drawing Partner Analysis App

The analysis platform supports: statistical comparison between participant groups, visualization of collaborative interaction patterns, interaction trajectory analysis, communication mode analysis, turn-taking behavior analysis, interaction timing analysis, and longitudinal co-creative behavior modeling.

For example, researchers can compare: novice versus expert collaborators, human-human versus human-AI collaboration, different interaction conditions, or different AI interaction strategies.

When statistically significant differences are detected between datasets, the system highlights the relevant interaction variables directly within the analysis tables. This allows researchers to identify which interaction dynamics contribute most strongly to collaborative creativity and participatory engagement.

The platform transforms co-creative interaction into a measurable and computationally analyzable phenomenon, helping bridge: computational creativity, HCI, enactive cognition, participatory sense-making, and interaction-centered AI research.

Creative Sense-Making Analysis App

The Creative Sense-Making (CSM) Analysis App extends the quantification framework by modeling collaboration as a continuously evolving interaction trajectory through time.

Users input Creative Sense-Making curves derived from one or multiple co-creative sessions. The system then visualizes the interaction dynamics using techniques inspired by stock market trend analysis and temporal signal modeling.

The analysis framework identifies periods where participants are: executing creative actions, waiting or observing, regulating interaction flow, adapting to collaborators, or transitioning between interaction states.

Different interaction states are visualized using color-coded dynamics:

Red represents active creative execution,
Yellow represents waiting or observational states,
Blue represents interaction regulation and coordination.

Try the Creative Sense-Making Analysis App

These evolving trajectories allow researchers to visualize how collaborative meaning and creativity emerge dynamically during interaction itself rather than only through final outputs.

The CSM Analysis App further supports: interaction trend analysis, participatory regulation modeling, adaptive coordination visualization, collaborative rhythm analysis, creative trajectory comparison, and temporal interaction pattern analysis.

Together, the AI Drawing Partner and its companion analysis systems establish a broader framework for: quantified co-creation, explainable co-creative AI, interaction-centered intelligence, and computational modeling of participatory interaction dynamics.

Rather than treating creativity as an isolated internal process, the framework positions creativity as an emergent phenomenon arising through sustained interaction between humans, AI systems, and evolving collaborative environments. (arxiv.org)

Why This Matters Today

The rapid rise of generative AI systems — including large language models, copilots, diffusion systems, and multimodal creative agents — has dramatically expanded public interest in human-AI collaboration. However, most modern AI systems still primarily focus on autonomous generation, output optimization, prompt-response interaction, and isolated artifact production. Within many contemporary generative AI paradigms, interaction itself often remains secondary to the production of outputs. Human users frequently function as prompt authors, selectors, editors, or evaluators rather than sustained collaborative participants within evolving creative systems.

The AI Drawing Partner explores a fundamentally different paradigm.

Rather than treating intelligence as isolated autonomous generation, the AI Drawing Partner investigates how creativity, meaning, and intelligence emerge dynamically through sustained participatory interaction between humans and AI systems. The platform positions co-creation itself — including turn-taking, improvisation, adaptive coordination, communication, and collaborative trajectories — as the primary object of analysis.

This interaction-centered perspective became increasingly important as generative AI systems evolved toward hybrid intelligence, collaborative AI, explainable AI, and human-centered AI systems.

Long before the current generative AI wave, the Drawing Apprentice and later AI Drawing Partner research lineage explored real-time human-AI improvisation, shared creative workspaces, participatory sense-making, quantified interaction dynamics, and explainable co-creative processes.

Unlike many modern black-box generative systems, the AI Drawing Partner was specifically designed to capture interaction traces, model collaborative trajectories, visualize co-creative dynamics, and analyze how collaboration unfolds through time.

The system therefore serves not only as a co-creative drawing application, but as an experimental research platform for studying interaction-centered intelligence, participatory creativity, adaptive collaboration, and quantified co-creation.

As AI systems become increasingly integrated into creative, educational, scientific, and professional environments, understanding how humans and AI systems collaborate, how interaction shapes intelligence, and how co-creative dynamics evolve through time may become just as important as improving generation quality itself. The AI Drawing Partner represents an early and influential step toward this broader interaction-centered paradigm of human-AI collaboration.

Explore Related Pages:

Page updated

Google Sites

Report abuse