Navies - useless in great power conflict
My entry to the First Sea Lord's essay contest sinks without trace...
My entry for the Royal Navy’s annual essay writing contest was always going to be an outside bet. I argued that outside of its role in nuclear deterrence, the fleet isn’t much use. But then, to show willing, I suggested that AI might soon improve matters. It wasn’t enough…. Anyway, here it is. Si vis pacem, para bellum, dark blue friends!
How can navies be used to dissuade rivals and deter opponents in an era of increasing geopolitical competition?
This is an essay in two parts. In the first I argue that navies are little use at deterring peer rivals in the modern era. In the second, I suggest this is changing.
But first, the premise of the question is agreeably contentious. Is this really an era of increasing geopolitical competition? By comparison to what?
There’s always been geopolitical competition, frequently intense, sometimes violent, and often with a maritime dimension. Russia has recently invaded Ukraine, throwing our recent national experience with ‘low intensity’ warfare into sharp relief. But when Britain’s army fought Russia in an earlier Crimean conflict, it lost over 20,000 men – fully a quarter of its modern strength. Construction of Britain’s new aircraft carriers cost £6bn. Not cheap, certainly, but still only a quarter of one percent of GDP. In 1628, when, for lack of ballast, Sweden’s Vasa went to the bottom on her maiden voyage, she took with her five percent of the Kingdom’s annual wealth. High stakes, amidst intense geopolitical competition. Perhaps, then, that’s the first lesson: even in an era of rapid technological change, some perspective helps. Today, military history is in retreat across university campuses and, whisper it, in military academies too.
A more recent comparison comes from Cold War debates about the utility of the fleet. The lessons are sobering – navies were not important in deterring actions by great powers, except in safeguarding the weapon of last resort aboard SSBNs. The record elsewhere is thin. In the Cuban missile crisis, blockade, an ancient tactic, demonstrated American resolve and signalled the President’s determination not to recklessly escalate. Soon afterwards, off Vietnam, the US Navy mounted a sustained bombing campaign from the decks of its flattops, rotating through the South China Sea. That campaign, however, failed to persuade Communists to cease infiltrating the South. Later, in the Falklands Conflict, the Royal Navy sailed half a world away to safely deliver its amphibious assault force, albeit having failed beforehand to deter invasion by a small, regional power.
Clearly, then, there was some utility from having big navies in the era of nuclear-armed superpowers. But very rarely in deterring or dissuading as part of great power competition. Too often, the fleet didn’t weigh in the balance in the top table. What would such a navy do in the event of direct hostilities between the USSR and NATO? A tactical nuclear weapon would destroy any task force. Perhaps, if hostilities started, the fleet could occupy a rung or two on the ladder of escalation – signalling by its sacrifice that matters were getting serious. Still, faced with the sobering costs of developing its own carriers, Khrushchev’s USSR demurred. After all, who needs a blue water fleet when you’ve a rocket force and vast reaches of steppe in which to hide it?
Alas, the problem remains vexing. Today, aircraft carriers deliver flexible power and convey intent. But they are scarce, expensive and vulnerable to increasingly ubiquitous and capable anti-ship missiles. Certainly, they’re a useful capability for expeditionary warfare against inferior adversaries like Iraq in 2003, or ISIS more recently; but they were far from essential in either case. In more intense competition against more capable adversaries, their utility is questionable.
Anyway, if it comes to a CCP assault on Taiwan, missiles and land forces will determine the outcome, not defending ships. As in earlier decades, carriers are simply tripwires that might trigger larger exchanges. Cheaper hostages are available. Meanwhile Russia, now denuded by its Ukraine misadventure, is no more a maritime power than ever it was.
New model navy
No paper is likely win the First Sea Lord’s prize by decrying both the surface fleet and the uniqueness of the contemporary challenges facing it. Happily for this author, however, there are some large, technologically driven changes on their way. These will change much in naval warfare, and they might even alter the deterrence calculus.
The developments I’m envisaging might not please traditionalists, especially those favouring carrier strike as the cornerstone of the fleet. But then the Royal Navy has long had an admirably mature approach to tradition and a capacity to reinvent itself for new times - as happened with the carrier itself, and before that when steam and steel supplanted sail and wood.
A caveat – it’s usually a mistake to see technology in isolation from the societies that both innovate and instrumentalise it. It matters greatly by whom these technologies are employed. And another caveat: the core of coercion, whether in deterring or compelling, remains psychological. Technologies, and the ways they are employed, can certainly factor. But ultimately, dissuading someone is a matter of mind, not material. How determined are they, and how determined are you? How far do you understand one another? There are no easy formulas.
The changes I’m contemplating flow from one technology type - artificial intelligence. In time, these will change the equipment, concepts and people involved in naval warfare, perhaps even challenging the very notion of an independent naval service. Among the many implications are those for dissuasion.
AI is a decision-making tool. That’s fundamentally what intelligence means: the ability to respond usefully to events. Today’s AI is increasingly capable - certainly of tactical activities, like manoeuvre, fire control and logistics. And tactical AI is scalable - so favouring distribution, saturation, and disposability. Uncrewed vessels, distributed shoals, drones launching drones: we are in a wacky-races, working through the possibilities.
Deterrence is more than tactics. But so too is AI. It’s becoming adept at more sophisticated reasoning - as demonstrated by large transformer models working in the medium of natural language, like the GPT family. Operational art may soon follow tactics along the road to autonomy. In ‘mosaic’ warfare, the autonomous unit in contact calls for reinforcement. Soon, an autonomous admiral might respond, knitting contacts together into a battle, eventually even into campaigns.
And AI strategists? That’s sacrilege for practitioners and concerned democrats alike. Time and space both permit human strategists to stay in the loop - it still takes twenty minutes for an SLBM to reach its target. But AI will increasingly be constructing the future First Sea Lord’s worldview, even if she ultimately decides. She will have grown up saturated by interaction with AIs - conversing with them, being tutored by them. Perhaps her education will feature immersive wargaming against Meta’s Cicero. That AI has recently attained expert-level performance in Diplomacy, a complex strategy game requiring both cooperation and deception, and an ability to persuade humans via natural language.
Meanwhile, her fleet, the distributed, robotic navy of the future, has no obvious centre of gravity against which to retaliate. It won’t be a speed bump on the road to general war, like the legacy fleets. Crucially, it might also offer more of a threat to enemy land and air forces. Scores of submersible missile boats lack the versatility of destroyers, but they’ll dispatch dense clouds of land attack missiles; and if many are sunk in the counterattack, so be it - there’ll soon be more arriving on station.
Clausewitz thought strategy most like a game of cards. But AIs can now defeat expert humans at poker by calculating the optimal hand. Schelling thought it more like a busy traffic intersection, the drivers jockeying for position - risky, intensely psychological, and far harder to calculate. Schelling is definitely a strategist for the coming AI age. A shoal of AI-powered missile boats; an ocean saturated with intelligent sensors hunting them; an admiral and her political masters deeply enmeshed within a powerful human-machine team. To me, all this makes an adversary less knowable and so predictable. It complicates any escalation ladder, giving more options for credible responses to provocations. From that incalculability, hopefully, comes greater stability - the central paradox of strategy. It might also bring the Royal Navy a renewed importance; for the first time since the advent of nuclear weapons, navies will possess conventional deterrent capability in great power competition.