That argument is flawed. You can destroy each other, so what? Everybody loses. We know that some countries have their weapons ready to go at the will of a single person or very few. What if that leader becomes crazy or incapable of reasoning? Your subs you't protect you in that scenario.

There are so many other ways you can prevent such attacks. Diplomacy, commerce, keeping good relations, etc. The threat of an axe on your neck makes it hard to look like you have good intentions in the first place.

That strategy demonstrably worked -- we got through the Cold War.

The problem is you can't always assume that the other guy has good intentions. If they're rational, then mutually assured destruction keeps everyone rational. If they're crazy, then that certainly is a problem, which is why there is enormous effort put into preventing proliferation and containing the capabilities of countries like North Korea and Iran.

You think mutually assured destruction is going to keep an Islamic extremist group rational?

Islamic extremists are not the only security threat to the USA. We continue to exist in a relativel tense peace with China, Russia, and other nation-states.

The ability to drop a bunch of nukes on someone on short notice is not to deter terrorists, it's to deter nation states - anywhere between North Korea through Pakistan.

The existence of nukes and the submarines that carry them is almost entirely orthogonal to the (relatively new) thread of terrorism.

No, that's what the NSA and CIA is for. :P

I realize it's a circular argument, but MAD is for other nuclear nations (or future nuclear nations).

Because the threat of force is the only real deterrent for some sorts of people / threats. Wishing it was different doesn't make it so.

That's why it was called M.A.D. - Mutual Assured Destruction - with dark humor. Nobody chose it, the world just evolved that way until it became codified.

In his book, "Star Wars," Lt. Bowman makes the case that the Strategic Defence Initiative (and similar missile defence systems) are bad ideas specifically because they disrupt MAD. His argument is that as long as we KNOW we are not being under attack, then we can all sleep at night. There is a point to this.

MAD wasn't just about deterrence, but also about ensuring mutual access to early warning information. The combination of the two is actually not a bad thing. "I know you can destroy me if I attack you, and I know you can see me if I start to do so" is not a bad set of factors. Unfortunately things like anti-satellite missiles and weapons in space really endanger that because if a small piece of space junk collides with a satellite over the other guy's territory, you may not know for sure that this is not the prelude to an attack.

So the actual MAD strategies as they developed through the Ford years were actually quite sophisticated and stabilizing. But what happens when your threat is a nuclear weapon in a twenty-foot container instead of an ICBM (because some country like North Korea knows we can have 747's with lasers shoot down launching missiles if we launch a first strike)?