Origin
From Formel to Form.
The Weltformel sought the single equation. WELT FORM starts from another question: what happens when an answer looks reliable before it has been verified?
Machines produce numbers, forecasts, decisions, and content ever faster. We build tools to put them to the test: check them against the real, stress their assumptions, explore their consequences, keep judgment contestable.
Who’s behind it
Signed judgment needs verifiable people.
Founder
Giuseppe Aceto
Designer, philosopher, and AI ethicist. He founded Relatronica in Zurich, a speculative and critical design lab on AI futures. Author of the Toolbox for Ethical Futures. His work sits at the intersection of algorithmic transparency, contestability, and human judgment, the ground WELT FORM is built on.
Our method
We ask you to put the model to the test.
Calibration
Estimates are checked against real outcomes. Uncertainty stays explicit.
Stress
Assumptions are perturbed to find where a system loses stability.
Simulation
We explore alternative trajectories and second-order consequences while action is still open.
Contestability
Reasons, hypotheses, and steps in the reasoning must be inspectable and attackable.
Where we come from
The research runs through design, AI ethics, decision theory, and computational simulation. Different disciplines, one question.
We keep them in conversation to measure the distance between output and verification.
How much can we trust an output, and what must we know before we act on it?
We publish research, record outcomes, and prefer verifiable models to systems that ask to be believed.