Science demo
Design an experiment that could produce a negative
Plan a test capable of returning the result that would change your mind.
The challenge
Experiments are designed to confirm, so a negative result is often not even possible.
The scenario
A hypothesis is agreed and the design must be capable of disproving it.
What goes in
- Hypothesis
- The claim being tested, with falsification criteria.
What you can ask
Illustrative prompts. Results depend on your own data and are not deterministic.
- “Design an experiment that could disprove this.”
- “What controls does this need?”
- “What would a null result look like?”
How Clearception approaches it
How the work is structured. Each step shows what its output is grounded in.
Read the criteria
SourceTake the falsification criteria as the design target.
Design for failure
AI-inferredBuild a test capable of producing the disproving result.
Specify controls
RecommendationState the controls the design requires.
What you get
The shape of what comes back — not a promised result.
Experiment design
RecommendationA design with controls, capable of a negative result.
Why it matters
- Informative either way
- Both outcomes teach something.
Related demos
Science
Turn observations into a testable hypothesis
Move from observation to a claim that could actually be wrong, with falsification stated.
4 minute walkthrough
- Hypothesis Design
- Fact / Inference Separation
- Evidence Trail
Science
State what would prove you wrong
Make the conditions for being wrong explicit before the data arrives.
3 minute walkthrough
- Hypothesis Design
- Fact / Inference Separation
Robotics
Plan a task, bounded by the safety plane
Turn a described goal into a task plan that the deterministic safety plane must accept before it runs.
4 minute walkthrough
- Deterministic Safety Plane
- Deterministic Policy Engine
- Evidence Trail
Want to try this with your own work?
Open Science and bring your own data. No signup is needed to read these demos.