Rachel Ostrand

Google

Rachel Ostrand is a Quantitative User Experience Researcher at Google in New York City. Her research focuses on language production and comprehension: in dialogue, when interacting with machines, and how people adapt characteristics of their language use based on the current context.

One area of focus is linguistic interaction between humans and computers (AI and large language models, chatbots, conversational assistants) and when and how people adapt their language production and expectations differently when talking with a machine as compared to another human, as well as what linguistic properties people expect and prefer in AI-generated language. As part of this, she studies partner-specific linguistic alignment, looking at the situations in which people modify properties of their speech to match those of their individual conversational partner (human or computer), as opposed to speaking with the same linguistic properties across multiple partners. She also investigates spontaneous speech as a remote, low-cost, low-burden diagnostic marker for detection of early-stage dementia and cognitive decline, including developing automated assessment metrics for detecting small cognitive changes through language production.