In a groundbreaking study, Large Language Model (LLM) ChatGPT-4 outshines human creativity by generating a higher volume of better and more cost-effective ideas than human creators.

Creativity is often regarded as an elusive concept, challenging to pin down but unmistakable when it appears. Psychologists have long attempted to measure creativity through various tools and tests, such as the alternative uses and Torrance tests.

However, recent developments in artificial intelligence are pushing the boundaries of our understanding of creativity. Enter GPT-4, a state-of-the-art Large Language Model (LLM) that has demonstrated remarkable creative prowess by surpassing human performance on these creativity tests.

Ethan Mollick, an advocate of artificial intelligence and a member of the Wharton School at the University of Pennsylvania, conducted a groundbreaking experiment that shed light on the potential of LLMs like GPT-4 to revolutionize idea generation and innovation. This experiment involved 200 students and GPT-4, both tasked with generating ideas for a product aimed at American college students with a retail price below $50.

Disclosure: This post may contain affiliate links, meaning we get a commission if you decide to make a purchase through links on this site, at no cost to you.

87.5% of Top Ideas in Creative Productivity Study Generated by ChatGPT-4

The results of this innovative study were nothing short of astounding. Of the top 40 business ideas evaluated by the judges, an overwhelming 35 were generated by ChatGPT-4. This exceptional performance by the AI model surpassed the creative abilities of the human participants.

Even more impressively, from a business perspective, the purchase intent for ideas produced by the AI was significantly higher compared to those generated by humans. This highlights the potential for AI-generated ideas to not only excel in creativity but also in market appeal.

Recommended reading: 10 Ways AI is Unlocking Creativity

One of the study’s most revealing aspects was its economic insight. The researchers found that a professional working in collaboration with ChatGPT-4 could generate approximately 800 ideas per hour, a staggering rate. In contrast, a human working alone, without the assistance of an LLM, generated a mere 20 ideas per hour.

When considering the fully loaded cost of a skilled professional, estimated at $500 per hour, the cost per idea generated by ChatGPT-4 was an astonishingly low $0.63. In contrast, a human working alone incurred a cost of approximately $25 per idea.

In essence, a human utilizing ChatGPT-4 for idea generation was approximately 40 times more productive than working alone.

Furthermore, the researchers highlighted the cost-effectiveness of using ChatGPT-4 for idea generation. At the time of the study, the API fee (application programming interface) for 800 ideas generated by ChatGPT-4 was approximately $20.

This stark contrast in cost efficiency underlines the potential for businesses to leverage AI-driven ideation as a strategic advantage in their innovation processes.

Study Highlights Great Opportunity for Human Creatives to Collaborate with AI

In summary, the study “Ideas are Dimes a Dozen: Large Language Models for Idea Generation in Innovation” reveals that ChatGPT-4, a leading LLM, outperformed human participants in creativity and generated ideas more quickly and cost-effectively. The implications of these findings are far-reaching, suggesting that businesses can harness the power of AI to enhance their innovation processes, drive creativity, and ultimately gain a competitive edge in the ever-evolving marketplace.

As Ethan Mollick aptly says, “We are running out of creativity tests that AIs cannot ace.” It seems that in innovation, the future belongs to the machines.

More helpful resources: