E MERGING T ECHNOLOGIES Partnering with AI: Intelligent writing assistance and instructed language learning

Robert Godwin-Jones

Abstract:In recent years, advances in artificial intelligence (AI) have led to significantly improved, or in some cases, completely new digital tools for writing. Systems for writing assessment and assistance based on automated writing evaluation (AWE) have been available for some time. That is the case for machine translation as well. More recent are synchronous feedback tools, such as Grammarly. That tool incorporates, as do others, predictive text technology, supplying automated sentence completion. Emerging writing assistance goes further, generating an entire text in response to a brief prompt. That capacity, along with significantly improved performance of both automated feedback systems and machine translation, is enabled through advances in AI, built on ever larger datasets and deep machine learning. While they differ in interface, functionality, and target audience, the available and emerging set of intelligent writing tools can be used to help learners improve the quality of their written texts. However, their use in instructional language learning has in some cases been controversial. In this column, we will be examining AI-enabled writing tools, reviewing the findings from research studies, and discussing their use in instructional settings. When integrated into writing instruction and practice, these digital tools have been found to offer significant benefits to both students and teachers. Teacher mediation aids learners in becoming informed consumers of language technology, as well as helping them to gain meta-linguistic knowledge. For researchers, intelligent writing tool use is optimally analyzed from a broad ecological perspective that examines the dynamic interplay of learner, software, and instructional environment.

Education,Linguistics,Computer Science