Co-Adaptation in an Ecosystem of Human–Machine Dyadic Interaction Introduction to the Forum

Abstract

In August 2025, The Washington Post published a headline that captured widespread public attention: “It’s happening: People are starting to talk like ChatGPT” (Aleksic, 2025). Reporting on findings by Juzek and Ward (2025), the article described how lexical items overrepresented in ChatGPT’s output such as delve were appearing with increasing frequency in human speech. Striking as this discovery may be, it is not entirely surprising. Artificial intelligence (AI) systems have rapidly permeated everyday life, and large language models (LLMs) such as ChatGPT are now used by millions for writing, learning, planning, and problem-solving. Many users engage with these systems not once, but repeatedly over time.