Pulling away from the mean.
LLMs, brand voice, algorithmically-boosted content: they're all converging toward the same center, the demographic average. A stereotype smoothed into something inoffensive and largely useless. The mean.
Before We Speak is built on a different premise. Real people don't experience the world from some neutral perch, they experience it through a particular reality. A reality shaped by where they grew up, what they believe, who raised them, what decade formed them.
Most AI knows the label but doesn't know the voice.
I built a taxonomy: country, leaning, region, culture, religion, generation. A working-class Black Catholic Gen Z in Atlanta doesn't experience the world the same way a secular white Boomer in rural Ohio does. The axes describe the reality, not the label.
Every briefing is shaped by real people. Contributors who actually live inside those identities are interviewed, and their responses ground what the AI produces. The voice comes from them, not from a guess.
The result is a platform that helps you understand the world someone else is inhabiting, not their politics...their actual world.
The taxonomy is still filling in. If your corner of it is missing, get in touch.
About me
I'm Peter Surrena, a product designer based in Brooklyn. My career has sat at the intersection of editorial, product, and technology, leading design and technology at Digiday Media, building competitive, behavioral, and market intelligence platforms at IDC, and now modernizing internal platforms at Bank of America.
Before We Speak started as a side project trying to reconcile two things. The first was watching media bubbles form and harden from inside the publishing world. The same event framed five different ways depending on the source. The second, the everyday experience of trying to talk with people I love who see things differently, and watching those conversations collapse even when everyone's acting in good faith.
The taxonomy approach came out of my love for patterns.