Conjecture · Criticism · Code
Popper Labs is an AI product lab and venture studio building systems for better problem solving. We use criticism to turn real problems into better explanations, then build the products, protocols, companies, and knowledge those explanations make possible.
“All life is problem solving.”Karl Popper
Popper is an agentic intelligence system personalized to your team’s real problems, unique ideas, and working knowledge. It turns conversation and live work into durable problem-solving context, so AI can reason and act from evidence instead of guessing from a blank chat or piecing together fragments across disconnected tools.
Problem
AI chat is useful, but its work often disappears inside the thread. Problems, decisions, and ideas still forming get lost before agents can use them and build on them effectively.
Building
A reimagined chat assistant that draws out the real tensions, remembers what matters, and turns conversation into durable context for autonomous problem solving.
Early access alpha
Conversations become problems agents can act on.
Problem
Hard problems quickly outgrow a chat thread. The reasoning behind important decisions, your best ideas, and your most curious questions rarely gets preserved or made explicit.
Building
A powerful workspace and agentic knowledge graph for your hardest problems, best ideas, and deepest explanations, compounding and evolving with you and your pursuits.
Early access alpha
Agentic AI that thinks with you, not for you.
Problem
Teams lose momentum when current truth gets scattered across chats, docs, calls, calendars, code, tickets, and forgotten updates. When action is needed, ownership and evidence are unclear.
Building
A communications and observability platform where teams and agents coordinate status, schedules, calls, evidence, and handoffs from the same current truth, without constant narration.
Beta
Shared truth becomes coordinated action.
Also built by Popper Labs
How we think shapes what we build. We treat every product, investment thesis, and organizational decision as a conjecture — something to be tested through criticism, not defended through consensus. The best ideas aren't the ones we protect. They're the ones that survive.
Progress starts with a real problem: the gap between how things are and how they could be. We let solutions follow.
Predictions are cheap without an explanation of why they should hold. We look for ideas that are hard to vary and clear enough to criticize.
Criticism is not a threat to good ideas. It is how good ideas survive, improve, and become reliable enough to build on.
A shared vocabulary for the ideas behind our work — problems, conjectures, explanations, and criticisms.
We don't optimize for headcount. We optimize for explanatory reach per person. Everyone here operates at the intersection of building and thinking — shipping products while developing the ideas behind them.
If you want to build AI products where the hard part is not just execution, but finding the real problem, making the reasoning explicit, and improving the system under criticism, we should talk.
careers@popperlabs.com →