Marissa Mayer's new AI assistant, Dazzle, takes a different route from the recent wave of personal agents. Instead of mining email, calendars, and shopping history, it builds a profile from a single source: the user's camera roll. The startup, which raised an $8 million seed round last December, argues that photos contain richer signals about hobbies, relationships, and preferences than text-based data.
In a hands-on demo, the assistant showed it could infer past travel from photos and suggest Mediterranean vacation spots, or recommend local activities like a pottery studio and a bioluminescent kayak tour. But it also had blind spots, such as failing to remember that the user's daughter already knew how to roller-skate. Mayer says the product improves with use and that analyzing photos feels less invasive than granting an AI access to emails and messages.
The approach builds on Mayer's earlier startup Sunshine, whose photo-sharing tool Shine was criticized and eventually shut down, though Mayer says it produced useful intellectual property. Dazzle is not yet as broadly capable as competitors like Meta's Muse or Instinct, but it offers a glimpse of an assistant that tries to know who you are rather than simply execute tasks.