A visitor from the telescope team asks the board a sharp question. "Why did Sky Sorter call my snapshot a comet?"
Rocket shrugs. "Because Sky Sorter said so?"
Raven shakes her head. "That is not an answer. We can do better."
Nova lights up Sky Sorter's version 3 decision tree on the wall. "Follow the path," she says. "Streak? No. Size 7 or more? Yes. Comet."
The visitor nods. "So I can see how it decided. And comet means I should watch for it again tomorrow."
Rocket looks impressed. "So how it works and what it means are two different questions."
Transparency is how much information about an AI system and its outputs is available to the people using it.
That matters even when people do not know they are using AI.
Accountability means someone is responsible for the system. Accountability depends on transparency.
Explainability is about how a system works. Interpretability is about what its output means for its job.
| Word | Question it answers | Sky Sorter example |
|---|---|---|
| Transparency | What can users find out about it? | A sign says Sky Sorter is a paper classifier made by the crew. |
| Accountability | Who is responsible? | Raven signs off on every new version. |
| Explainability | How does it work? | The decision tree path shows each step. |
| Interpretability | What does the output mean for the job? | Comet means: watch for it again tomorrow. |
| Statement | True or false? |
|---|---|
| Accountability depends on transparency. | ? |
| Transparency only matters if people know they are using AI. | ? |
| Explainability is about how a system works. | ? |
| "Because the AI said so" is a full explanation. | ? |
Clear thinking. Tomorrow the board runs its full review in the Explorer Lab.