There is considerable debate about whether LLMs, which are based on probability of word occurrences across an enormous corpus of text, can yield ideas that are creative. This study provides some clues.
The research involved participants writing a micro-story, which was defined as an eight-sentence fictional story. While the study focused on short fiction, the authors point out that short-form creative writing is also used in the business world in roles such as marketing. In the training world, content creators are frequently called on to write scenarios that illustrate a concept, video scripts, or engaging summaries of content that serve to drive course enrollment. So there is reason to believe that this finding might be relevant for training developers as well.
It’s important to note that the impact of AI on creative writing tasks is not uniform. AI tends to bring the performance of team members who are less creative closer to the level of those who demonstrate high creativity. It doesn’t hurt highly creative performers. It just doesn’t help them increase their creativity much.
The finding that those using AI may produce results that are more similar to each other compared to writers not using AI may be a feature more than a bug for training use cases. Training content is not typically rated based on novelty alone. Consistency can create a better experience for learners taking multiple courses.
