The Good Tech Companies - Reimagining Memory How Autograph Turns Phone Interviews into a Timeless AI Portrait

Episode Date: June 18, 2025

This story was originally published on HackerNoon at: https://hackernoon.com/reimagining-memory-how-autograph-turns-phone-interviews-into-a-timeless-ai-portrait. Autograph turns weekly phone calls into AI-generated digital legacies, preserving voices, memories, and emotions for generations to come. Check more stories related to machine-learning at: https://hackernoon.com/c/machine-learning. You can also check exclusive content about #ai-memory-preservation, #autograph-ai, #digital-legacy-tools, #ai-voice-clone-ethics, #emotional-ai-storytelling, #cristian-cibils-bernardes, #future-of-memory, #good-company, and more. This story was written by: @jonstojanjournalist. Learn more about this writer by checking @jonstojanjournalist's about page, and for more stories, please visit hackernoon.com. Cristian Cibils Bernardes’ startup Autograph captures weekly phone interviews to build lifelike AI replicas that preserve a person's voice, stories, and emotional memory. Built on trust and privacy, Autograph is redefining legacy, enabling users to simulate future conversations and leave timeless, meaningful messages for loved ones.

This audio is presented by Hacker Noon, where anyone can learn anything about any technology. Reimagining Memory How Autograph Turns Phone Interviews into a Timeless AI Portrait, by John Stoyan Journalist Christian Sibyls-Bernards, founder of the early stage company Autograph, ice-driven by a vision that merges deep technological innovation with profoundly human intent. Born in Paraguay and trained at Stanford in symbolic systems, a discipline at the intersection of computer science, philosophy, and cognitive psychology, Sybill's-Bernards is no stranger to complexity. But the problem he's tackling with autograph is surprisingly simple. How do we preserve and pass on human memory? There's an asymmetry

in time, says Sybill's-Bernards. The older generation has accumulated a wealth of lived experience and now has free time. But they're often lonely, meanwhile, the younger generation is busy building their lives and doesn't always have time to sit and listen. We build a bridge between them, a time vault of memories. Autograph does this by recording weekly phone interviews with users via Walter, its AI historian. These conversations serve as raw material for remarkable output. An AI-generated digital replica that sounds like the user, speaks like the user, and even remembers what the user remembers. The eye signature of a life at its core, Autograph is not about replicating humans but about

preserving their essence. The process begins with transcripts of the phone interviews. These are indexed by memory, identifying people, places, and key experiences, and then used to train a fine-tuned AI model that mimics the user's vocal patterns and expressions. "'We fine-tune a model so that it speaks like you andises your same turns of phrase,' Sibyls Bernard says. The magic sauce I.S. mostly in how we do the memories and how we convince the AI that it's playing a character named William, for example. While terms like, clone, might be tempting, Sybils Bernard's deliberately avoids them. We call them autographs, he says. Because they are your signature. You should trust it as if you were signing with it. The result is an

impressively faithful digital presence, one that can respond to factual questions about Apperson's life, offer consistent opinions, and preserve emotional nuance for years to come. Simulations, messages, and emotional continuity beyond memory preservation, Autograph opens doors to experiences that we re-previously the domain of science fiction. The AI replicas can simulate future conversations, personal decisions, and even playful what-if scenarios. You can simulate a weekend, a career decision, or a difficult conversation, says Sybill's Bernards. Or for fun, simulate yourself living in middle-age Rome and see how quickly you learn Latin. More than entertainment, the emotional applications are profound. Messages recorded today can be delivered years later.

A user could leave a message of encouragement, advice, or love to be accessed long after they're gone. You don't know when you'll need an, I'm proud of you, or, I love you, he shares. This tool lets you be there, even decades after you're gone. Designing for trust, not just technology, Autograph is intentionally built as a data trustee rather than a data harvester. Unlike major tech platforms that leverage personal information for scale, Autograph is grounded in trust and longevity.

It's a weekly phone call with a friend, essentially, Sibyls Bernard says. Over time, we build a profile that includes not only your stories but your photos, contacts, and social footprints. It's built from the ground up to be your memory bank. As development continues, autograph is being rolled out deliberately. A full autograph can achieve up to 80% fidelity within a month of regular conversations, but the long tail, those intimate memories and rare moments, takes time to capture. That's where the subscription model comes in, ensuring the AI evolves as life

unfolds. The future. Infinite you looking ahead. Sibyls Bernardz envisions a future where digital replicas aren't mere archives but active participants in our lives. A digital replica has the awesome benefit that you don't get just one, you get infinite, he explains. A version of you for every task, for every conversation, for every simulation. Ultimately, the purpose extends beyond productivity. It becomes a tool for meaning making. What is the most meaningful life I can live? Simulate me the most meaningful life, Sibyls Bernardes muses. Sometimes I wonder if I'm not the autograph already, trying to figure it all out.

Through autograph, Sibyls Bernards is not just rethinking memory, he is redefining digital legacy. His work invites us to imagine a world where human experience is never truly lost, but instead evolves, informs, and supports the ones we love. To learn more about how AI is changing how we store our collective memory, connect with Christian Sibyls Bernards on LinkedIn or X. Thank you for listening to this Hacker Noon story, read by Artificial Intelligence. Visit HackerNoon.com to read, write, learn and publish.