Pulp Corporation · EST. 2026 · Brooklyn, New York
We help humans understand humans.
Pulp is a research company in Brooklyn. We measure how people talk, argue, and make up their minds, build tools for people whose work runs on conversation, and publish what we find.
Photo: José Martín Ramírez Carrasco
Why it matters
Technology doesn’t make you a better thinker.
It never has. Habits do. Every new tool changes what people practice, and AI is no different.
AI doesn’t care about people. People do. Some people, anyway. That’s who we build for: people who want to argue better, listen longer, and change their minds for the right reasons.
Walking away solves nothing.
We’d rather do the work than sit it out: make it clearer what people actually said, what they meant by it, and what the machine did with it.
Three ways we look at people.
Hover or tap a plate to see what it measures.
Large groups over time: how views move across a city, a country, a decade.
One person at one moment: who they are, what they care about, how they’re feeling, and where they are.
Why people decide what they decide, written as numbers for machines and plain words for people.
What they said.
What they meant.
What you can hold us to.
- Quality
- We check things more than once before you see them.
- Reliability
- What works today works tomorrow. Updates don’t break what you already use.
- Value
- More out of the conversations you’re already having, using less compute.
- Data literacy
- We show our sources and say how sure we are.
- Verified people
- We can confirm a real person took part without exposing who they are.
- Credible references
- When we cite media or public figures, you can see where it came from.
What we make, in three.
The Laboratory
Pulp101, Thinkwell, typodojo, Suade, and Loosethreads: everything we make and test.
The platform
The server SDK: Radiance, Windfall, TrueAura, and Groundwater, plus MCP.
The Foundation
A nonprofit, 501(c)(3) pending, preparing an American workforce for work after AI.
Research
See us in Chicago this November.
We’re presenting at the American Society of Criminology’s annual meeting. Our work was also accepted at the American Sociological Association, and we’ve given seminars at the University of Florida. More is coming soon.
Data stories
- From Stone to ScreenComing soon
- How people ask, and what it costs to understand themComing soon
- Understand first: how Pulp cuts the electricity AI usesComing soon
Everyone talks
at once.
From one person to a whole public.
Reflection
Look back at what you said, and what you meant.
One to one
Keep track of who promised what.
One to many
Find out how a message actually landed, from real people.
Many to many
Run discussions where the better argument wins.
Photo: Vladyslav Tobolenko
Tell us what you’re trying to understand.
A question, a dataset, an argument that keeps going in circles. We’ll tell you plainly if we can help.