Dangerous to whom? is indeed the question. When we were designing orbital strategic defense systems, we were working in a mission timeline window of about 300 seconds. That's five minutes. If the system did not receive the go-ahead in time, it would miss all opportunities to intercept boosting ICBMs. We concluded that it had to be autonomous by default with maybe a veto switch commanded by the ground. It would be like a house or a factory with an automatic fire-extinguishing system that would not activate unless given permission.
Most strategic weapons are like robot dogs with nuclear bombs on their collars instead of brandy kegs. "Go, kill!" "Yeth, thir!" We don't worry so much that they will go off on their own, but that our men might not follow orders to send them forth.
Dangerous to whom? is indeed the question. When we were designing orbital strategic defense systems, we were working in a mission timeline window of about 300 seconds. That's five minutes. If the system did not receive the go-ahead in time, it would miss all opportunities to intercept boosting ICBMs. We concluded that it had to be autonomous by default with maybe a veto switch commanded by the ground. It would be like a house or a factory with an automatic fire-extinguishing system that would not activate unless given permission.
Most strategic weapons are like robot dogs with nuclear bombs on their collars instead of brandy kegs. "Go, kill!" "Yeth, thir!" We don't worry so much that they will go off on their own, but that our men might not follow orders to send them forth.