Finetuning Artificial Intelligence for Entrepreneurial Pitching Effectiveness

The rise of artificial intelligence (AI) presents challenges yet also opportunities to understand whether and how it can be deployed by entrepreneurs to tell better stories to attract resources. In this study, we first introduced the notion of AI-generated versus human-generated storytelling and explored the potential merit of LLM AI to enhance entrepreneur’s pitching effectiveness. In doing so, we fine-tuned ChatGPT GPT 4 that was fed with winning pitches from Shark Tank and had it transformed and refined losing pitches and evaluated their differences. We conducted three evaluations that involved two human experts to score the AI-refined versus original (losing) pitches using measures of pitch effectiveness, quantitative natural language analyses of the two types of pitches, and qualitative feedback from entrepreneurs and specialists (n=17) to make sense of the results. Overall, we found that AI-refined pitches generally outperformed their original human-generated (losing) pitches while also offering some caveats to the findings. We discuss what LLM AI will mean for entrepreneurship storytelling research and its implications for theory and practice and propose future research agendas. © 2025 Academy of Management. All rights reserved.