Can Large Language Models become Innovation Experts? : Emulating Expert Evaluation of Innovation in Startups with Large Language Models

Abstract

Startups are drivers of innovation and economic growth, but the uncertainty associated with startups makes the evaluation of their innovativeness difficult. Currently, startups are mostly evaluated by experts (e.g., domain experts, potential investors, researchers). However, experts are scarce and can only evaluate a limited number of startups due to their limited cognitive resources as well as time and cost constraints. Large language models (LLMs) might be a suitable and scalable alternative for expert evaluation of innovation in startups due to their emergent abilities. Once LLMs are trained, they can be applied to diverse tasks and contexts without additional training. This master’s thesis aims to answer the question of how LLMs can be used to evaluate the innovativeness of technologies and strategies presented in startup pitches by applying prompt engineering techniques and fine-tuning LLMs.