What makes people think that AI systems are creative? New research shows that it depends on their perception of creative behavior. These findings have implications for how we research and design creative AI systems, and they also raise fundamental questions about how we view the creativity of others.
‘AI plays an increasingly larger role in creative practice. Whether this means we should call it creativity is a different question. ” Niki Pennanenthe main author of the study. Pennanen is studying AI systems at Aalto University and has a background in psychology. He conducted experiments with Aalto and other researchers at the University of Helsinki to find out whether people think robots would be more creative if they saw more creative behaviors.
In this study, participants were initially asked to evaluate the robot’s creativity based on only the still life maps they made. They were told that the robot is powered by AI, but in fact, the robot has been programmed to replicate drawings commissioned by researchers from artists. This deception allows people to measure how people think about creativity without the need for robots to be creative, which will introduce too much variability between drawings.
Next, the study participants evaluated the creativity of drawings when they not only see the final product, but also the video of the drawing process – these lines appear on the page instead of the robot creating them. In the final stage, participants can see all three elements when scored: final product, process and robot production drawings.
The results show that drawings are considered more creative, just as revealing more elements of creative behavior. “The more people you see, the more creative they think it is.” Christian Guckelsberger, Assistant Professor of Creative Technology of Aalto and Senior Author of the study. As far as I know, we are the first to study the impact of perceived products, processes and manufacturers in a separate and controlled way, not only in the context of AI, but more broadly. ”
The power of perception
Understanding how people evaluate the creativity of robots or other artificial systems is important to think about how to design them – but it is not clear what the appropriate design choice is. “This study shows that revealing more information about the process and producers may help people think of the creativity of the system,” Guckelsberger said. “But if we add elements to make the AI system look more creative, even if the system actually performs the same, we might question that it is actually a good thing.” “In some cases, this may help – for example, it may be a way to help people keep co-creating systems. But in other cases, it may give people a deceptive impression of the creativity of artificial systems.
“Our discovery helps resolve this conflict by giving us a better understanding of our own human biases. This study makes them more transparent, and from a user’s point of view it is also important for us to understand how the design of a system affects our perception of it. ” Guckelsberger said.
In addition to these social and design implications, these findings are also of great significance to the study of creative AI systems. If our judgment of creativity depends on how the system is proposed, future research should control for this factor. Similarly, based on these findings, existing research needs to be re-evaluated – comparing the creativity of different systems without considering their presentation differences may lead to false conclusions.
Another interesting question raised by this study is that it tells us about ourselves. “Now we’ve found out what people think of AI creativity… Does this apply to what people think of others, too?” asked Guckelsberger.
Is shape important?
The researchers also conducted experiments using two different robot designs. Their goal is to test whether people score differently on creativity based on the shape of the robot, as earlier work proposes a link between shape and perceived creativity.
The team tested whether people saw different levels of creativity when the still life was drawn by a stylish arm-like robot or a more mechanical plotter robot. Keeping the robot consistent with the drawings from one participant to another is challenging. “I think our biggest difficulty is the physical robot itself. We did a lot of work in robotics and drawing processes to keep everything the same so we could make scientific and rigorous comparisons. ”
The researchers were surprised to find that there was no significant difference in the ratings of the two robots. They plan future work to further study this counterintuitive outcome, and other factors that influence our perception of creativity. “We are interested in doing more research on which bias affects our assessment of creative and embodied AI systems and how these effects occur,” Pennanen said.
These findings, along with other forms of artistic and creative expression, should also be confirmed. To make it easier for others to replicate their work and base it on it, researchers followed a strict open science practice. As artificial systems become commonplace, understanding the factors that shape our perception of their creativity is crucial for effective design – and may also shed light on how we identify human creativity.
The preprinted version of the paper is now available at the following website: The final publication will be available on April 7 EET from the same link.
If our report has been informed or inspired, please consider donating. No matter how big or small, every contribution allows us to continue to provide accurate, engaging and trustworthy scientific and medical news. Independent news takes time, energy and resources – your support ensures that we can continue to reveal the stories that matter most to you.
Join us to make knowledge accessible and impactful. Thank you for standing with us!
