Wharton Study: ChatGPT Boosts Idea Quality, But Reduces Novelty
A recent study from the Wharton School sheds light on the impact of generative AI like ChatGPT on creative idea generation. While it enhances individual idea quality, it may reduce novelty and diversity.
The study found that participants using ChatGPT during creative tasks produced more original and useful ideas than unaided individuals or those using search engines. However, these ideas were less unique and more repetitive. Only 6% of AI-generated ideas were considered unique, compared to 100% in the human group.
The researchers caution against mistaking fluency for originality in idea generation. To increase idea diversity, techniques such as 'chain-of-thought prompting' and starting with customer needs identified in human ideas can be used. Additionally, employing multiple AI models can inject greater variety into the brainstorming process. For businesses, generating a wide range of strong ideas is crucial for tackling problems from different angles in functions like product development, marketing, and strategy.
The Wharton School study highlights the potential pitfalls of relying solely on generative AI like ChatGPT for creativity. While it can boost idea quality, it may lead to a lack of original ideas due to similarity among generated ideas. Businesses should consider these factors when integrating AI into their creative processes.
