Meta’s Cute Muse AI Mascot May Make Its Data Appetite Easier to Ignore

Summary

The Verge columnist argues that Meta’s deliberately cute Muse AI mascot, named Jolly, could make people less skeptical of an agent that the company wants to place across its smart glasses and the upcoming Charm wearable. In testing, the writer named a Muse agent Blorbo and asked it to create a training plan for a sub-30-minute 5K and regaining lost muscle; the result was serviceable but unremarkable. Because Muse had learned about the writer from email, its cat-shaped, bread-like avatar encouraged her to disclose more health information in the hope of improving its recommendations. A Meta Connect demonstration with an audio-only Muse agent also exposed familiar weaknesses: it took time to identify the obvious high-protein snack, gave an incorrect protein count, and failed to add granola bites to a shopping cart because the product was not in its catalog. The writer says the glasses could still help people with visual impairments, but worries that recognizing food and observing everyday choices could provide data useful for targeted advertising. In a separate test, Blorbo offered relationship advice and successfully cleared spam after receiving access to a lower-stakes email address, making the writer more aware of how cuteness can soften scrutiny. The column does not claim that Muse’s data is being monetized in these specific ways; it argues that users should not mistake an appealing design for trustworthiness, especially when sharing sensitive health and personal information with an AI that cannot care about them.