AI news ·
How Our Perception Shapes the Creativity We See in AI and Beyond
People judge AI creativity based on how much of its creative process they see. Revealing more steps boosts perceived creativity but raises ethical concerns about true capability.
Is AI Truly Creative? Creativity Depends on What You See
What makes us consider an AI system creative? New research suggests that our judgment depends largely on how much of the creative act we witness. This insight could influence how creative AI systems are designed and how we perceive creativity more broadly.
AI is increasingly involved in creative fields, but whether we label it "creative" is still debated. Researchers at Aalto University and the University of Helsinki explored whether people rate a robot as more creative when they see more of its creative process.
How the Study Was Conducted
Participants first rated the creativity of robots based only on still life drawings. They were told the robots used AI, but the drawings were actually reproductions of an artist’s work programmed into the robot. This approach allowed the study to focus on perception rather than the actual creativity of the robot.
Next, participants saw videos showing the drawing process—the lines appearing on the page—but not the robot itself. Finally, they viewed all three elements: the final drawing, the drawing process, and the robot creating the image. The results were clear: the more participants saw, the more creative they judged the robot’s work to be.
The Role of Perception
Understanding how people judge AI creativity is key to designing these systems. Revealing more about the process and the creator can boost perceived creativity. But adding elements to make an AI appear more creative without improving its actual performance raises ethical questions.
This could be useful in keeping people engaged with co-creative AI systems. Yet it might also mislead users about the AI’s true creative capability. Recognizing these biases helps us design more transparent systems and better understand how presentation shapes our judgments.
These findings also impact research on creative AI. Studies comparing creativity across systems should control how these systems are presented, or risk drawing inaccurate conclusions.
What This Means for Human Creativity
The research opens a curious question: if perception heavily influences how we judge AI creativity, might it also affect how we see creativity in other people? This calls for further exploration into the biases that shape our evaluation of creativity overall.
Does the Robot’s Shape Influence Creativity Judgments?
The study also tested whether the robot’s appearance affected creativity ratings. Participants saw drawings made by either a sleek, arm-like robot or a more mechanical plotter robot. Despite expectations, there was no significant difference in how creative the two robots were perceived.
Keeping the drawing process identical across different robots was challenging but crucial for a fair comparison. The results suggest that robot shape might not strongly influence creativity judgments, though further research is needed.
Future Directions
- Investigating what other factors influence our perception of creative AI.
- Confirming findings across different artistic styles and creative domains.
- Exploring biases in evaluating embodied AI systems.
The researchers followed strict open science practices to enable others to replicate and build on their work. As AI systems become more common in creative fields, understanding how presentation affects perceived creativity will help build better interactions—and could also shed light on how we recognize creativity in humans.
A pre-print of the paper is available here. The final version will be accessible from the same link on April 7 at 15:30 EET.