The Former Yahoo CEO Just Built an AI That Reads Your Entire Camera Roll and It Knows Things About You That Google Doesn't
TL;DR: Marissa Mayer's new startup Dazzle launched in September 2026, offering an AI assistant that analyzes your entire photo library to surface personalized recommendations and complete everyday tasks. Backed by $8 million in seed funding, Dazzle works entirely on-device and promises Google-level photo intelligence without sharing your data. This is the most personal AI product I've seen yet.
I've been covering AI announcements for years, but this one stopped me cold. Marissa Mayer — the former Google VP and Yahoo CEO who's been quietly working on something since 2022 — just launched an AI assistant called Dazzle, and it does something I didn't think was possible on a phone: it reads every photo you've ever taken and turns that history into a working model of who you actually are.
Not who you told Google you are. Not who your purchase history says you are. Who you actually are, based on where you've been, what you've eaten, who you hang out with, and what makes you stop and take a picture.
What Dazzle Actually Does
Dazzle has two operating modes, and both of them are built on the same foundation: a complete analysis of your camera roll. The first mode handles immediate, task-based requests — things like "find me a restaurant similar to the one I went to for my sister's birthday" or "show me all the receipts from my last business trip." It scans your photos in real time and surfaces exactly what you need, often without you being able to articulate what you're looking for.
The second mode is more unsettling and, frankly, more impressive. It runs in the background, builds a model of your patterns and preferences, and proactively makes recommendations based on what it knows. If you photograph coffee every morning, it learns your routine. If your photos cluster around a specific neighborhood on weekends, it starts suggesting things to do there. It connects dots across months of your life that you've already forgotten about.
The Privacy Angle You Need to Know
Here's where Mayer has made a genuinely interesting bet. Dazzle does all of this processing on-device. The AI model runs locally on your iPhone — no photos are uploaded to a server, no image data leaves your phone. The app uses Apple's Neural Engine and on-device machine learning frameworks to analyze your library without any of it touching Dazzle's cloud infrastructure.
This is a real differentiator. Google Photos is exceptional at understanding your photos, but the trade is that your images live on Google's servers where they train their models. Dazzle is betting that a meaningful segment of users want that intelligence without the trade. Based on the waitlist numbers Mayer's team shared at launch — over 400,000 sign-ups in the first 48 hours — that bet looks correct.
The Funding and the Team
Dazzle raised $8 million in seed funding in December 2025 from a group that includes Sequoia Capital and early investors from Mayer's previous ventures. The company has been operating in stealth for nearly three years, which explains how a first-time founder-CEO with a Y Combinator pedigree managed to build something this polished as a v1.
Mayer's team at Dazzle includes former members of Google's Photos team (which she originally built during her tenure there) and iOS engineers who worked on Apple's Photos app. That combination — the people who built the best photo AI in the world and the people who built the best photo UX on mobile — is not a coincidence. This is a deliberate product bet from people who know exactly where the edges of the technology are.
What Google and Apple Don't Do That Dazzle Does
Google Photos will find your dog across 10 years of photos. Apple Photos has Memories that surface moments you forgot. Neither of them will look at a photo of a restaurant menu from two years ago and cross-reference it with a photo of a receipt from last month to remind you that you've been meaning to go back.
That's the layer Dazzle operates at — not just retrieval, but inference. The system builds a persistent model of your preferences that updates as you take new photos. It's closer to having a personal assistant who has been paying attention to your life than it is to a search tool.
Should You Try It
The waitlist is open now and Mayer's team is rolling out access in batches. I've requested access and I'll share a hands-on review as soon as I have it. The on-device claim needs to be verified independently, and I want to understand exactly what data, if any, leaves the device for personalization updates. But the concept is the most compelling personal AI product I've seen in the current wave — and the timing, just as Apple Intelligence rolls out its own expanded photo features, makes the competitive dynamics genuinely interesting to watch.
What's your experience? Drop a comment below! 👇
Would you trust an AI that reads your entire camera roll if you knew it stayed on-device? Or does the idea of any app analyzing 10 years of your photos feel like too much?
#MarissaMayer #Dazzle #AIAssistant #ArtificialIntelligence #Privacy #TechNews #StartupAI #PersonalAI #CameraRoll #AI2026
Comments
Post a Comment