New study challenges belief that AI can't handle creative tasks ...
Cryptoasset regulation remains fragmented, with emerging markets and developing economies lagging behind in global push, as revealed in a new study conducted by Cambridge Judge Business School.
Explore our degree programmes and non-degree programmes to discover scholarship and loan opportunities. Additionally, delve into our Executive Education portfolio consisting of over 40 meticulously crafted programmes designed to enhance your skills and comprehension in alignment with your organizational, personal development, and career objectives.
Insights from the Study
According to conventional wisdom, artificial intelligence (AI) and specifically Large Language Models (LLMs) were believed to pose a threat to human tasks that are more mundane and repetitive, with human creativity holding a significant advantage. However, previous research on the creativity aspect has yielded inconsistent and varied results.
The latest study, co-authored at Cambridge Judge Business School, addresses this inconsistency by comparing the collective creativity of LLMs to that of groups of humans across 13 creative tasks spanning different domains such as divergent thinking, problem-solving, and creative writing.
Key Findings
The study found that when an LLM is questioned multiple times, its collective creativity is equivalent to that of 8-10 humans. Additionally, the study suggests that with more responses requested, 2 additional LLM responses equate to one extra human. This challenges the notion that AI is incapable of handling creative tasks and indicates that LLMs, when optimally utilized, may compete with a small group of humans in future work scenarios.
Implications and Future Outlook
Although LLMs ranked on average in the 46th percentile against human participants across all tasks, the study underscores the potential of LLMs to contribute significantly to collective creativity in various settings. As LLM capabilities continue to advance, the study serves as a crucial milestone in highlighting the comparable performance of LLMs to humans in creative tasks.
Ultimately, the study suggests that organizations can leverage the collective creativity of LLMs, especially in scenarios involving limited human participation, while acknowledging that the most creative ideas may still emanate from human input.
For further details, you can access the full study here.
