Voice preservation means either archiving actual recordings of someone's voice or using AI to generate new speech in their vocal pattern based on text you provide. The technology lets you hold onto something that words alone cannot capture—the irreplaceable sound of them—though it also raises hard questions about the difference between preserving what was and creating something that wasn't.
Voice preservation using AI is capturing someone's actual voice—through recordings, conversations, messages—and preserving it in a way that survives them. Some platforms go further and use voice synthesis (AI-generated audio) to extend that voice beyond what was originally recorded. You might have 3 minutes of recorded voicemails from your grandmother, but using voice synthesis, you could have more—perhaps readings of family recipes in her voice, or birthday messages for grandkids not yet born.
This is ethically complex and deeply personal. But for many grievers, hearing a loved one's voice again is profoundly healing—it's one of the first things people miss after loss.
Voicemails, voice memos your person left, interviews you record with them, phone call recordings (if legal in your jurisdiction), videos with clear audio. Some tools like 11Labs Voice AI can learn from small samples of voice and recreate new audio in that person's voice pattern—their accent, their tone, their speech rhythm.
Grief attacks through the senses. You reach for the phone to call them before you remember. You think you hear them in a crowd. Hearing their actual voice—not a memory, not a reconstruction, but the real sound—can ground you. It can also help combat the slow fade where, months or years later, you can't quite remember what their voice sounded like. Preservation fights that erosion.
For some grievers, voice synthesis offers something different: continuation. If your parent died before they could record a birthday message for your wedding, or before they could tell you they're proud, voice synthesis (used respectfully, with family consent) can offer closure that voice preservation alone can't.
This technology requires consent and intention. Using someone's voice to say things they wouldn't say is impersonation and disrespectful. Using it to say things they would say (reading their own writing, continuing their values) feels different. Most families using this have explicit agreements about what the preserved voice can be used for.
Start with preservation: collect any existing voice recordings, create new ones if the person is still living (interview them), transcribe anything important. Store these securely (you don't want them lost to a failed hard drive). Then decide: is preservation enough, or do you want to work with a voice synthesis tool? If the latter, use it intentionally and tell family members what you've done.
Try this: If you have one voicemail or voice recording from someone you've lost, listen to it this week. Write down what you notice—not just the words, but the tone, the cadence, the feeling it evokes. That's the irreplaceable quality of voice. Consider: Is there a way to preserve this, officially, so you don't lose it to technology degradation? Even just saving it to multiple secure locations is a form of preservation.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.