Dazzle bets your camera roll knows you better than your inbox
Former Yahoo CEO Marissa Mayer unveiled Dazzle this month, a personal AI assistant backed by an $8 million seed round that builds its understanding of you from a single source: your photos. The launch lands amid a wave of similar tools from Meta and others, but Mayer argues the camera roll offers richer context than email, calendars, or shopping histories ever could.
"I think that photos are an underappreciated source of information," Mayer said. "You'll be surprised what we can learn about you and how good a job we can do with your photos."
Dazzle splits its functionality into two experiences. For immediate tasks, it scans recent photos to pull details - populating your calendar from an event flyer or finding a repair person after spotting a broken garage door. Second, it mines your photo library to generate personalized ideas for vacations, gifts, and activities.
What the camera roll reveals
Mayer's pitch hinges on a simple premise: if a photo is worth a thousand words, a camera roll is worth millions. By analyzing images stored on your phone, Dazzle claims to understand hobbies, interests, food and style preferences, how you spend your time, and who you spend it with.
"We understand whether or not you like to ski, where your most recent trip was, what types of things your kids are into," Mayer said.
Mayer's previous startup, Sunshine, launched an AI-powered photo-sharing tool called Shine in 2024. That product drew criticism for its outdated design, failed to attract widespread usage, and eventually shut down. Mayer said the work nevertheless produced "interesting IP" that informed Dazzle's development.
Testing the assistant
Early demos show both promise and blind spots. When asked for vacation recommendations, Dazzle suggested Mediterranean destinations, likely picking up on past trips to Spain and Greece. It also recommended Sicily, a place visited four years ago. But it failed to remember that a daughter already knows how to roller-skate when asked about birthday gift ideas.
Despite the hiccups, the suggestions felt personal. Dazzle recommended a local pottery studio and a bioluminescent kayak tour in Tomales Bay - ideas that would not surface from an assistant that only reads calendar entries and email threads.
Mayer uses Dazzle herself to brainstorm family outings. The assistant deduced from her photos that her family enjoys escape rooms and suggested several San Francisco Bay Area locations she had not heard of.
Privacy as a differentiator
Mayer contends that users may feel more comfortable handing over photo libraries to an AI than granting access to sensitive data like emails and messages, particularly given security concerns raised about competing products such as Meta's Muse and Instinct.
She said Dazzle prioritizes privacy and discards any personal information the AI flags as sensitive. The assistant can be accessed through its app or via text.
For professionals working with AI tools, understanding how different assistants build user context matters. Those exploring Generative AI Courses or AI Agent Courses will encounter similar design choices about data sources and personalization across the tools they evaluate.
Why this matters for product developers and writers
Dazzle's approach raises a question worth sitting with: what data source actually produces the most useful personalization? If the camera roll outperforms email and calendar data for understanding preferences, product teams should reconsider which signals they treat as authoritative. For writers and marketers, Dazzle's photo-driven recommendations hint at a future where AI-generated content reflects lived experience rather than inbox history - a shift that could change how personalization copy is written and how creative briefs are built. The product is not broadly useful yet, but the signal is clear: context sources will determine which assistants feel genuinely personal.
Your membership also unlocks:
