Documenting decades of family through pictures has been made easier these days. We have Facebook and other social media apps to store even the photos from the '90s or '80s. In the age of AI, whoami.wiki creator Jeremy discovered a cupboard filled with 1,351 loose photographs of his family.
Instead of simply organizing them, he transformed the experience into an ambitious AI-driven memory project.
Turning Family Photos Into a Living Encyclopedia
While sorting through the images with his grandmother, Jeremy uncovered forgotten memories that she had not recalled in years. And because he's inspired to showcase it, he built a Wikipedia-style archive using MediaWiki, the same platform that powers Wikipedia.
Within just two evenings, he recreated his grandmother's wedding as a structured encyclopedia entry: complete with an infobox, scanned photos, captions, and contextual links to historical events of that time.
Read more: OpenAI Eyes Gigawatt-Scale Fusion Power From Helion as Sam Altman Steps Down Amid Deal Talks
Claude AI Decodes Travel and Life Data
Jeremy later expanded the experiment using Claude AI by Anthropic. He tested it on 625 vacation photos from Coorg, India, where no GPS metadata was available. Despite this, the AI inferred locations, transport routes, and travel context using visual clues alone.
The system's capabilities grew even more impressive with a larger dataset from a 2022 trip to Mexico City. With 291 photos, 343 videos, and supporting data such as Google Maps history, Uber receipts, bank transactions, and Shazam logs, Claude cross-referenced everything to rebuild a detailed narrative of the trip.
What's more amazing is that it even matched Ticketmaster records with video timestamps to identify soccer matches and used music recognition data to determine songs playing in specific moments.
Reconstructing a Digital Life History
According to Boingboing, Jeremy pushed the project further by feeding more than 100,000 messages from Facebook, Instagram, and WhatsApp, along with thousands of voice notes spanning over a decade. The AI mapped relationship patterns, emotional shifts, and forgotten personal milestones hidden in digital archives.
He noted that the system didn't just organize data. Instead, it also reshaped how he understood memory itself, revealing how interconnected personal experiences become when analyzed at scale.
Private, Open-Source Memory System
Unlike cloud-based AI tools, Jeremy's system is fully local and open source, ensuring that all personal data remains private and under user control.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.
