The study, published in the Journal of Creative Behaviour, analyzed large language models and determined that their creative output cannot exceed a score of 0.25 on a normalized scale from 0 to 1. This puts their creative capacity at the level of an average human and precludes them from matching the originality of highly creative professionals in fields such as writing, music, art, and scriptwriting.
According to Cropley, generative AI can convincingly mimic creative processes but fundamentally lacks the ability to generate truly original, effective works. He stated, "While AI can mimic creative behaviour - quite convincingly at times - its actual creative capacity is capped at the level of an average human and can never reach professional or expert standards under current design principles."
Cropley's research highlights a common misconception about creativity in AI. Many assume that the ability of models to produce stories, images, or poems equates to creativity. However, he clarifies, "Generating something is not the same as being creative. LLMs are trained on a vast amount of existing content. They respond to prompts based on what they have learned, producing outputs that are expected and unsurprising."
He goes on to explain that a significant proportion of people view AI systems as creative due to differences in individual creative ability. "Typically, 60% of people are below average when it comes to creativity, so it's inevitable that a sizeable slice of society will think that LLMs like ChatGPT are creative, when they're not. Highly creative people will recognise the weaknesses in the generative AI systems," Cropley noted.
The study is the first formal calculation of AI creative capacity based on internal LLM mechanics. Cropley asserts that, while generative AI may be a useful tool in creative industries, it cannot replace individuals with exceptional creativity. "A skilled writer, artist or designer can occasionally produce something truly original and effective. An LLM never will. It will always produce something average, and if industries rely too heavily on it, they will end up with formulaic, repetitive work."
He adds that true expert-level AI creativity would require new computational architecture capable of generating ideas that depart from patterns established by existing data. The study offers reassurance to creative professionals concerned about automation's impact, concluding, "This research shows that the world still needs creative humans - perhaps more than ever."
Research Report:The Cat Sat on the ... ? Why Generative AI has Limited Creativity
Related Links
University of South Australia
All about the robots on Earth and beyond!
| Subscribe Free To Our Daily Newsletters |
| Subscribe Free To Our Daily Newsletters |
