Year: 2024
Type: dissertation
Abstract: This thesis investigates the potential of Large Language Models (LLMs) to improve social media content creation, with a focus on LinkedIn posts in conjunction with the company Knowit Connectivity. As digital marketing increasingly relies on artificial intelligence (AI) to personalize user encounters, it is vital to understand the capabilities and limitations of LLMs. This study investigates if LLMs can match or outperform humans in creating engaging, contextually relevant LinkedIn postings. Using Google's Vertex AI for GDPR compliance and robust fine-tuning capabilities, this study refines AI outputs using approaches such as Reinforcement Learning from Human Feedback (RLHF). The study compares the engagement metrics and user perception of LLM-generated posts created by humans and AI using a combination of surveys and NLP analysis. Initial studies reveal that, while LLMs excel at producing technically correct and contextually appropriate content, their capacity to innovate and truly engage with consumers is still outweighed by human creativity. However, under some settings, fine-tuned LLMs have the potential to greatly alter social media strategy by personalizing material that is appealing to the target demographic. This thesis adds to the ongoing discussion about AI's role in strategic communication by presenting insights that could impact future digital marketing strategies and AI research.
Author: Tanudin, David
Institution: KTH Royal Institute of Technology
Language: English
Subfield: Marketing
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7064396533