Dazzle Builds Personal AI From Your Photos
Marissa Mayer's Dazzle uses camera-roll context as the foundation for a personal AI assistant.
Dazzle is testing a different thesis in the race to build personal AI: your camera roll may reveal more about your real life than your inbox.
What happened
The startup, led by former Yahoo CEO Marissa Mayer, unveiled a personal AI assistant that derives much of its context from photos stored on a user's phone.
Dazzle can extract information from images and use it to perform actions such as creating calendar events, identifying objects or finding services related to things captured in photos.
The company previously raised an $8 million seed round.
Why it matters
Most personal assistants build context from text-based sources such as email, calendars, messages and browsing history.
Photos contain a different kind of information: places visited, people met, products owned, documents photographed, hobbies and offline activities.
If an assistant can interpret that information accurately, it can potentially understand parts of a user's life that conventional productivity tools never see.
The bigger picture
Personal AI is becoming a competition over context.
The winning assistants may not simply have the best models. They may be the products that can securely access the richest, most useful representation of a user's life.
That makes data access, privacy and trust just as important as intelligence.