Evolution of Co-Creative AI

A Timeline of Research, Systems, and Frameworks

Over more than a decade of research, this work of Nicholas Davis PhD and collaborators, has explored how humans and AI systems collaborate creatively through interaction, participation, adaptive regulation, and co-creative sense-making. The image above marks the major publication and conceptual milestones of Davis et al. The central trajectory of this research can be summarized as a gradual shift from studying creativity as a property of individuals toward understanding creativity, cognition, and intelligence as phenomena that emerge through interaction. The first diagram (above) maps the contemporary Co-Creative AI research ecosystem. The second diagram (see bottom of the page) traces the historical evolution of the core ideas and publications that led to its development.

2013 — Co-Creativity Foundations

The earliest phase of this research explored creativity as an interactional and participatory process rather than an isolated act of individual production. Emerging from Human–Computer Interaction and computational creativity research, this work investigated how humans collaborate creatively through dynamic interaction, mutual influence, improvisation, and shared sense-making.

These early foundations established a long-term research direction centered on: interaction-centered creativity, collaborative cognition, distributed creative systems, and participatory human–AI collaboration.

This period laid the conceptual groundwork for later work in co-creative AI, quantified interaction dynamics, and enactive co-creation.

2014–2016 — Drawing Apprentice

This period introduced Drawing Apprentice, one of the early real-time co-creative AI drawing systems designed for collaborative artistic interaction between humans and AI systems.

Grounded in enactive cognition and improvisational interaction, the Drawing Apprentice explored how AI systems could function not merely as generation tools, but as active creative partners participating dynamically within shared artistic workflows.

The project investigated: collaborative drawing, reciprocal creative influence, interaction-centered AI, participatory sense-making, and co-creative artistic dialogue.

This work marked an important transition toward AI systems designed for: adaptive participation within human creative interaction.

2017 — Creative Sense-Making

Creative Sense-Making: Quantifying Interaction Dynamics in Co-Creation introduced a framework for modeling and quantifying the interaction dynamics unfolding during co-creative collaboration.

The work proposed that creativity could be studied not only through final artifacts, but through: interaction rhythms, cognitive fluctuations, communication dynamics, turn-taking behavior, and collaborative adaptation across time.

The paper introduced methods for: coding co-creative interaction, visualizing creative trajectories, and analyzing collaborative sense-making as an evolving temporal process.

This work became a foundational step toward: quantified co-creative AI systems, interaction-centered computational creativity, and co-creative interaction modeling.

2018–2021 — Quantified Co-Creation

This phase expanded Creative Sense-Making into broader frameworks for: quantified co-creative AI systems.

The research explored how co-creative interaction could be: computationally modeled, quantified, visualized, and analyzed dynamically during human–AI collaboration.

The work introduced: interaction traces, creative trajectories, collaborative state modeling, co-creative visualization systems, and quantified interaction frameworks for artificial creativity.

This period increasingly shifted focus toward: creativity as measurable interaction dynamics.

The AI Drawing Partner and related systems served as experimental platforms for investigating: collaborative adaptation, interactional flow, explainable co-creation, and dynamic human–AI participation.

2022–2024 — Explainable Co-Creative AI

This phase introduced frameworks for: explainable and transparent co-creative AI systems.

The research argued that artificial media should not remain a black box of hidden collaborative processes. Instead, co-creative systems should preserve and expose: interaction histories, participation dynamics, creative provenance,m and collaborative influence structures.

This work proposed that quantified co-creative systems could function as: explainable co-creative systems.

The resulting frameworks explored: creative provenance, explainable artificial media, adaptive collaboration modeling, and transparent human–AI interaction dynamics.

This period established a strong bridge between: co-creative AI, explainable AI, provenance systems, and participatory Human–AI Interaction.

Unlocking the Black Box of Artificial Media
Creative Provenance
Explainable Co-Creative AI
Quantified Interaction Frameworks

2025+ — Enactive AI

The current phase of research extends co-creative AI into broader frameworks for: Enactive AI.

This work investigates how AI systems sustain: adaptive participation, collaborative sense-making, interactional coherence, and regulation under dynamic drift across time.

Building on earlier work in quantified co-creation and participatory interaction, Enactive AI explores cognition as: an ongoing process of interaction and adaptive regulation.

Current frameworks investigate: interactional drift, participatory coherence, adaptive co-regulation, distributed cognition, and interaction-centered AI architectures.

The broader goal is to help establish foundations for AI systems capable of: meaningful collaboration, adaptive participation, and sustained co-creative interaction within dynamic environments.

Distributed Creativity (2011–2013)

This period explored creativity as a distributed process emerging across people, artifacts, and environments rather than solely within individual minds. The work established early foundations for thinking about human-computer creativity as a collaborative phenomenon rather than a property of isolated systems.

Artistic Computer Colleagues (2014–2015)

The introduction of enactive cognition transformed the research from creativity support tools toward genuinely collaborative AI systems. During this period, the concept of the Artistic Computer Colleague emerged, proposing that AI systems could participate in creative activity alongside humans rather than simply generating outputs.

Participatory Sense-Making (2015–2016)

The focus shifted from the individual participants to the interaction itself. Drawing on participatory sense-making, this work investigated how meaning emerges through ongoing engagement between humans and AI systems, laying foundations for interaction-centered approaches to co-creation.

Creative Sense-Making (2017–2019)

This phase introduced methods for modeling and quantifying interaction dynamics through concepts such as activity traces, creative trajectories, and sense-making curves. The central question became not merely whether a system was creative, but how creativity emerged through interaction over time.

Enactive AI & Human-AI Co-Creation (2020–Present)

The research expanded beyond computational creativity toward a broader theory of interaction-centered intelligence. Recent work investigates how cognition, creativity, and adaptive behavior emerge through ongoing interaction among humans, AI systems, and environments, leading to frameworks such as Human-AI Co-Creation, Interaction-Centered Intelligence, Enactive AI, and Enactive Drift Regulation.

Page updated

Google Sites

Report abuse