Intelligent Writing Assistants in the era of Large Language Models

Vipul Raheja discusses the evolution of intelligent writing assistants, driven by advances in large language models and human feedback. He highlights the shift from prescriptive systems to more collaborative ones, while also addressing their current limitations in quality, personalization, and usability. The talk delves into his research on developing intelligent and interactive writing assistants, focusing on text revision as an essential part of the writing process. He presents a taxonomy for text revision, computational modeling of iterative text revision using state-of-the-art techniques, and the development of human-in-the-loop text editing systems to improve interpretability and efficiency. The research explores system-driven, user-driven, and mixed-initiative approaches to writing assistance, demonstrating how human-machine collaboration leads to higher quality revised documents.