On the Creativity of Large Language Models

TL;DR

The paper investigates whether large language models are truly creative by applying Boden's criteria of value, novelty, and surprise and placing them within the Rhodes press and person framework. It analyzes LLM capabilities and limitations through historical context, theoretical lenses, and practical implications, arguing that current models exhibit utility and limited novelty but lack transformational creativity and intentionality. The authors discuss easy vs hard problems in machine creativity, emphasize the role of environment, motivation, and self-evaluation, and propose continual learning and domain adaptation as future directions. They also explore the societal and legal ramifications, advocating for human–AI co-creativity and responsible policy to harness benefits while mitigating risks.

Abstract

Large Language Models (LLMs) are revolutionizing several areas of Artificial Intelligence. One of the most remarkable applications is creative writing, e.g., poetry or storytelling: the generated outputs are often of astonishing quality. However, a natural question arises: can LLMs be really considered creative? In this article, we first analyze the development of LLMs under the lens of creativity theories, investigating the key open questions and challenges. In particular, we focus our discussion on the dimensions of value, novelty, and surprise as proposed by Margaret Boden in her work. Then, we consider different classic perspectives, namely product, process, press, and person. We discuss a set of ``easy'' and ``hard'' problems in machine creativity, presenting them in relation to LLMs. Finally, we examine the societal impact of these technologies with a particular focus on the creative industries, analyzing the opportunities offered, the challenges arising from them, and the potential associated risks, from both legal and ethical points of view.