Can an AI System Be Creative? A Critical Perspective from Art and Engineering

2026-07-22 · Source: Artificial Intelligence · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Human-Computer Interaction · Depth: Expert, quick

Summary

This paper critically analyzes whether artificial intelligence (AI) systems can be creative, integrating perspectives from electrical engineering, machine learning, and the arts. Drawing on Margaret Boden's framework of three creativity properties (novelty, surprise, and value) and three process types (combinatorial, exploratory, and transformational), the author argues that AI systems are structurally limited. While AI demonstrates strong capabilities in combinatorial creativity, it is significantly bounded in exploratory creativity and fundamentally incapable of transformational creativity. A central limitation identified is AI's lack of mechanisms for serendipity, accident, or the unexpected, and the absence of a "subject position" to recognize such chance events. The paper concludes by proposing a realistic and generative model for human-AI creative collaboration, exemplified by its own composition.

Key takeaway

For creative technologists exploring AI's artistic potential, recognize that current AI systems are best suited for combinatorial tasks. You should design collaborative workflows where human intuition and serendipity drive exploratory and transformational creative leaps. Focus your AI tools on generating variations or assisting with structured elements, reserving the subjective recognition of unexpected insights for human oversight to achieve truly novel and valuable outcomes.

Key insights

AI systems are fundamentally limited in true creativity, particularly transformational types, due to a lack of serendipity and subjective recognition.

Principles

• AI excels in combinatorial creativity.
• Exploratory creativity is bounded for AI.
• Transformational creativity is beyond current AI.

Method

The paper proposes a model for human-AI creative collaboration, illustrated through concrete experiments, and demonstrated by the paper's own composition via a deliberate human-AI collaborative process.

In practice

• Focus AI on combinatorial tasks.
• Design human-AI creative workflows.
• Integrate human serendipity into processes.

Topics

• AI Creativity
• Human-AI Collaboration
• Boden's Creativity Framework
• Combinatorial Creativity
• Transformational Creativity
• Serendipity in AI

Best for: Research Scientist, AI Scientist, Creative Technologist, AI Ethicist

Related on AIssential

• AI Won’t Replace Human Creativity. It Will Unleash It. — AI will transform creativity by overcoming execution bottlenecks, enabling easier realization of ideas.
• Under What Conditions Can a Machine Become Genuinely Creative? — Genuine machine creativity requires structural transformation of incomplete situations through recursive intervention, not just novel output.
• TapNow AI and the Rise of the Creative Canvas — The future of AI creativity lies in agentic creative canvases that enable structured, repeatable visual workflows, not just single outputs.
• Artificial Intelligence and Human Creativity: Partners or Rivals? — AI's role in creativity mirrors historical tool adoption, initially sparking fear but ultimately fostering new artistic expressions.
• AI and the New Renaissance of Art and Design — New technologies, like AI, transform creative fields by liberating human expression rather than destroying existing practices.
• Automated Creativity Evaluation of Language Models Across Open-Ended Tasks — The framework offers a scalable, domain-agnostic method to evaluate LLM creativity by separating measurement from task.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial Intelligence.