Human-AI Co-Creativity: Advances, Opportunities, and Challenges

TEXT START: This survey article has grown out of the human-AI co-creativity workshop organized by the authors at the ICML 2026 conference.

The Dissection

This is an agenda-setting and community-building document, not an economic analysis. It packages generative AI’s displacement of creative labor as a collaborative research field, emphasizing workshops, practitioners, and future directions. The term “co-creativity” preserves the image of the human as an enduring production partner while leaving ownership, control, bargaining power, and labor demand largely unexamined.

The text may describe real workflows in which humans and models jointly produce artifacts. That is operational cooperation, not proof of equal economic necessity.

The Core Fallacy

It conflates human participation with human indispensability.

A human may provide prompts, taste, curation, verification, domain context, or final approval while AI performs the expanding share of cognitive production. If those functions become cheaper to automate, standardize, or concentrate within AI systems, the human remains a servitor, liability shield, or aesthetic interface—not an independent economic sovereign.

“Co-” describes the workflow. It says nothing about who owns the models, captures the surplus, or retains access to employment. Under P1–P3, co-creativity can increase output while destroying the wage-and-consumption circuit that made creative participation economically widespread.

Hidden Assumptions

• Human creativity remains scarce and difficult to replicate.
• Human judgment will retain market power rather than become a low-paid verification layer.
• AI will remain a collaborator controlled by users, rather than capital controlled by Sovereigns.
• Research advances will translate into broad productive participation instead of concentrated rents.
• Technical, ethical, and social “challenges” are manageable without confronting ownership and coordination failure.
• Community-building can preserve human relevance after the labor market no longer requires most human contributors.
• Greater creative output benefits creators broadly rather than primarily benefiting the owners of compute, models, platforms, energy, and distribution.

None of these assumptions is established by the supplied abstract.

Social Function

Primarily ideological anesthetic, prestige signaling, and transition management—with a partial truth embedded inside.

The partial truth is that humans and AI can produce novel work together. The anesthetic is treating that fact as evidence that humans will remain economically central. The workshop gives institutions and professionals a respectable vocabulary for adapting to automation without stating the harder conclusion: collaboration can be the cultural form of displacement.

The Verdict

This abstract is not a rebuttal to the Discontinuity Thesis. It is a polished wrapper around the transition. It records researchers building a field around human-AI interaction while avoiding the decisive question of whether human labor remains necessary once AI dominates cognitive production. Co-creativity may preserve niches for Sovereigns and indispensable Servitors, but as a general social settlement it is hospice language for the mass creative worker.