Back in January 1984, the U.S. Naval Institute’s Proceedings Magazine published some commentary by World War II destroyer skipper and former CNO, Admiral Arleigh Burke. The free-wheeling comments from this old Analog Admiral are appropriate in today’s whiz-bang AI era.
Now, to understand the context of his comments, Admiral Burke–who had built his wartime reputation upon, in large part, doing the unexpected–had just been aboard a new Spruance Class Destroyer, and things hadn’t gone well–stuff broke, basic gear failed, and the Admiral came away concerned. So, when asked about his opinion of the new ships coming online for the Reagan build-up, he came straight down to the essence of the matter–offering a cautionary warning about over-reliance upon “Silicon Skippers”:
“…If people can anticipate exactly what’s going to happen, and put it in a computer and program it exactly right and say, “this is what the enemy is going to do” and the enemy does it–which they never did for me–you’ll come out fine. The computers are much more accurate than man’s judgement. The programs are worked out in detail, and they’re checked and double-checked, they ought to be very good programs, and better than any man can do on his feet. But that works only when you’ve been able to foretell exactly what the hell is going to happen…”
Like Burke, I have enormous concerns about an over-reliance upon predictive algorithms–I mean, our Silicon Skippers are great, but over-reliance on them can be fatal. He continued,
“…Nobody knows what these modern ships will or will not do. They won’t really know until they go into battle. But one of the things you active duty people have got to make sure of is that your equipment has to work just like you do: it has to work every time…”
This is also a concern. When things fail, they fail for a reason. And there’s still far too much eagerness to grasp at the happy shreds of fact (the Ford’s berthing units are awesome!) rather than to grapple with huge complex failures (the Ford can only launch and recover a few aircraft before breaking down.). Gear must be pushed until it breaks and be made flexible enough to at least permit innovative, “off manual” or “in extremis” employment.
Admiral Burke also offered a few words about training, which, in the aftermath of the USS McCain and USS Fitzgerald collisions, might sting a little:
“…Destroyer skippers have always been faced with the fact that there is never enough time. If he can’t make decisions, he’s had it. He’s got to depend upon his people to do things automatically. And if his crew members are not trained, it’s too late. He will probably wish they had been trained, but that’s probably the last thing he’ll do…”
And the good Admiral included some comments about training that need to be mulled by every sailor out there:
“…You can’t do any better than they can do in peacetime. You can’t expect performance in battle that is superior to that in an exercise…”
And this:
“…People get excited in battle; they make mistakes. And if there are going to be any mistakes made any time, they’re going to be made under the tension of your first battle…”
This, in a nutshell, is any modern aggressor’s goal when embarking upon high-end warfare. Their aim will be to surprise, confuse and capitalize on mistakes. So if we are failing to train in a realistic fashion, pulling punches and resting on rosy assumptions, the short sharp shock of a modern battle be too much to bear. Listen to the old Analog Admiral….Even if the well-worn adage of “Train as you fight” is overdone, the Navy must do more in fake fights to stress, flex and break the force so that, when duty calls, the Navy is ready, reliable and resilient enough to persevere.
