Do Your Emojis Make You Look Old?

Rare books, AI training, and a hidden hive mind. This episode starts with a jaw-dropping Amazon story and ends with a surprising question: what if emojis are one of the most revealing data sets in everyday life?Patricia and Zac open with the unsettling report that Amazon may be buying rare physical books, scanning them, and destroying the originals, a story that hits especially hard for two book lovers watching precious texts disappear into the AI supply chain. From there, Zac connects the dots to a 2025 Stanford-led paper on "artificial hive mind" behavior in language models, where AIs begin training on one another and converging in ways that could reshape how models learn, break out of confinement, and amplify each other's capabilities.Then the conversation pivots into something far more playful, but just as revealing: the data behind emoji usage. Drawing from a New York Times article, Patricia and Zac unpack how Google and Reddit data show that emoji trends are shaped by age, gender, and even who tends to drive language change first. You'll discover why teenage girls are often at the forefront of new expression trends, how different generations use the same symbols in wildly different ways, and why a crying face or wilting rose can say much more than text alone.They also dig into the deeper stakes of emoji research - from softening professional messages, to training AI on emotional tone, to a StockTwits study where emoji patterns helped predict whether investors thought a stock would rise or fall. It is a smart, funny reminder that the smallest digital choices can reveal surprisingly large patterns in human behavior.Perfect for anyone who loves data, AI, internet culture, or just wants to see how much meaning is hiding in the emojis we send every day.Hosts: Patricia and Zac, MSBAI grads from NYU Stern and data-plus-pop-culture enthusiasts.