Idea
Move the decision inside the deadline
Remote control requires a loop. The machine reports what happened, the operator interprets it, and a command travels back. If the entire event finishes before that loop can close, a skilled operator at the other end cannot intervene in time. Faster judgment cannot compensate for a message that has not arrived.
A Mars landing is such a case. The vehicle must respond to its sensors through an onboard sequence while people on Earth wait for evidence of the outcome. Human expertise moves upstream into design, simulation, and testing. The immediate action belongs to the system already at the scene.
A decision must live where it can arrive before the deadline.
Autonomy can be a consequence of distance before it is a matter of intelligence.
Where it breaks down
Local control does not mean unlimited discretion. A landing system operates within engineered rules and assumptions, and unexpected conditions can still defeat it. Latency also varies with planetary positions; there is no single permanent Earth–Mars communication time.
How it connects
A closed control loop cannot help when its messages arrive after the event ends.
This idea appeared in When help arrives too late, the Involves connection for September 21, 2026, which asked: What must a machine decide before its operator can hear about the problem?
Check yourself
Why did landing decisions need to occur aboard Curiosity?
Communication with Earth could not close the control loop in time. Exactly. The deadline was shorter than the useful remote response time.
What the sources establish
Curiosity's landing sequence had to operate onboard because Earth communication took longer than descent.
Sources
Challenges of Getting to Mars: Curiosity's Seven Minutes of Terror (NASA Science), Engineering video on entry, descent, landing, and communication delay
How Do I Land on Mars? (NASA Mars Exploration), Landing sequence and delayed Earth confirmation