Rocket climbs into the observation dome, where the station telescope points at a patch of dark sky.
"Every snapshot card starts right here," he says, patting the telescope's housing.
Raven checks the control panel. "The camera at the back catches the light. That is the sensor."
"So Sky Sorter can never see the sky directly," Rocket says slowly. "It only gets what the sensor sends."
Nova circles the dome, her lights reflecting off the glass. "Exactly," she says. "If the sensor misses something, the sorter misses it too."
Raven opens a fresh page in the mission log and writes a heading: Sensor Plan.
A sensor is a device that detects something from the world, such as light, sound, heat or touch.
What a sensor picks up is called a signal. A signal is raw information, before anyone works out what it means.
Computers perceive the world using sensors. Without sensors, a computer only knows what people type in.
| Sensor | Signal it detects | Example |
|---|---|---|
| Camera | Light | The telescope camera catches starlight |
| Microphone | Sound | A recorder picks up a voice |
| Thermometer | Temperature | A probe measures how warm the dome is |
| Touch screen | Touch | A screen detects where a finger presses |
| Statement | True or false? |
|---|---|
| A signal is raw information picked up by a sensor. | ? |
| Computers perceive the world using sensors. | ? |
| A sensor already knows what its signal means. | ? |
| If a sensor misses something, a sorter that depends on it can miss it too. | ? |
Getting meaning from sensor signals is hard. A camera records light, but it does not know it is looking at a comet.
Making computers "see" and "hear" well enough for practical use is one of AI's most significant achievements so far.
Even so, real AI does not see or hear the way you do. It works with signals and patterns, not experiences.
Good start. Tomorrow you will follow a signal all the way to meaning.