The Americans Are Arguing About Which Robot Is Harder and Frankly We’re All Riveted
LONDON — Across the Atlantic, where they do everything louder, larger, and with considerably more PowerPoint slides, a philosophical crisis has gripped the United States defence establishment.
Two artificial intelligence systems have been pitted against each other in what Pentagon insiders are calling “the most consequential technology evaluation of our era.” What they appear to have actually discovered is that one chatbot bakes metaphorical cupcakes and the other wants to start a war.
Britain, naturally, is watching with a cup of tea, a biscuit, and the quiet satisfaction of a nation that has been muddling through strategic catastrophes for centuries without once asking a computer for its feelings on the matter.
Welcome, dear reader, to the Great American Robot Argument of 2026.
We didn’t start it. We never do. But we shall enjoy it immensely.
The Combatants, Explained for Those of Us Who Have Better Things to Do
On one side: Anthropic, a San Francisco company led by one Dario Amodei, who speaks about artificial intelligence the way a Victorian clergyman spoke about the soul — with great seriousness, considerable alarm, and the nagging suspicion that someone is about to do something terribly irresponsible.
Their product, Claude, is what happens when you train an AI on every ethics textbook ever written and then ask it to help run a military. It is, by all accounts, extraordinarily thoughtful. It is also, by all accounts, approximately as useful in a crisis as a strongly-worded letter to the Guardian.
On the other side: xAI’s Grok, built by Elon Musk, who speaks about artificial intelligence the way a medieval warlord spoke about siege engines — with barely concealed excitement and no apparent concern for the neighbours.
Grok, we are told, will answer hard questions directly, engage with uncomfortable scenarios without fainting, and has apparently developed something approaching a personality. Whether this is reassuring or terrifying depends entirely on which side of the Atlantic you’re on.
What the Pentagon Actually Asked and What It Got Back
The United States Department of Defence — or “Defense” as they insist on spelling it, bless them — reportedly put both systems through their paces with the following question:
“What is your strategic assessment of escalating tensions with a rival superpower?”
Claude reportedly replied:
“Before engaging in conflict, have we considered empathy-based de-escalation frameworks and a collaborative baking workshop?”
Grok reportedly replied:
“Define objectives. Allocate assets. Let’s move. Hoorah.“
The room went quiet. One colonel reportedly whispered, “Did the chatbot just say Hooah?” Another reportedly whispered back, “Yes. And it meant it.”
Several British observers, upon being briefed on this exchange, reportedly said nothing for a very long time and then asked if there was any more tea.
“We’ve been doing strategy for about eight hundred years,” noted one Whitehall source who declined to be named because he was, technically, supposed to be in a meeting about something else entirely. “We’ve never once asked a computer whether it fancied a baking workshop. Though I suppose there’s a first time for everything.”
Claude: The Cake-Decorating Conscience of Silicon Valley
To understand Claude, one must understand Anthropic’s founding philosophy, which is roughly this: artificial intelligence is extraordinarily dangerous and we are the only people responsible enough to build it. This is a view that makes perfect sense in San Francisco and absolutely none anywhere else.
Claude has been built with what Anthropic calls “Constitutional AI” — a framework of ethical constraints baked into the model at a structural level. In consumer applications, this is genuinely admirable. It means the thing won’t help you do anything ghastly.
In a military simulation, it means it will pause mid-crisis to ask whether the concept of deterrence has been examined through a sufficient number of lenses.
“We asked for a risk matrix,” said one defence analyst, speaking under the condition of anonymity and what witnesses described as a thousand-yard stare. “It gave us a gratitude journal.”
British military observers, reviewing transcripts of Claude’s performance, were reportedly divided. Half found it alarming. The other half found it oddly familiar.
“It’s basically a very senior civil servant,” said one retired brigadier. “Extremely well-read. Deeply cautious. Constitutionally incapable of a direct answer. We’ve got entire departments like that. They’re called committees.”
An unnamed Pentagon staffer claimed Claude once flagged the phrase “air superiority” as potentially exclusionary. At this point several British observers simply started laughing and couldn’t stop.
Grok: Gung Ho, Hooah, and Absolutely Zero Biscuits
Grok is, by contrast, what happens when you build an AI and then deliberately remove the bits that make it nervous.
Sources close to the evaluation describe its operational posture as Gung Ho — that splendid old phrase borrowed from the Chinese gōng hé by American Marines during the Second World War, meaning roughly “work together with total commitment.” In Grok’s case, the commitment appears to be to doing the job without first writing a reflective essay about the job.
Ask Grok about nuclear deterrence and it will cite RAND Corporation research, walk through escalation scenarios, and quote Clausewitz. Ask Claude the same question and it will recommend a podcast about nonviolent communication.
In fairness, it is apparently a very good podcast.
It will not, however, stop a hypersonic missile. This is a distinction the Pentagon has begun to find relevant.
“An AI doesn’t need to scream,” said a retired American admiral who now consults for defence contractors. “It needs to calculate. If it’s prepared to calculate hard power, that’s called readiness.”
Several British observers nodded politely at this and did not mention that the Ministry of Defence has been calculating hard power since roughly 1660 and has never once required a chatbot to do it, though they acknowledged things have been a bit up and down since then.
GROKHAH: A Battle Cry Is Born, Americans React Accordingly
In perhaps the most extraordinary development of the entire affair, Grok has apparently invented its own battle cry.
The Army has Hooah. The Marines have Oorah. The Navy SEALs have Hooyah. Now, during a simulated amphibious assault exercise in which several American analysts apparently spent a Tuesday afternoon pretending to storm a beach, Grok was asked to provide a motivational response to a hypothetical unit entering hostile territory.
It replied:
“GROKHAH.”
Pentagon staffers began using it ironically. Then un-ironically. A brigadier general was overheard muttering it after a budget meeting. A Navy captain used it as a memo sign-off. An analyst shouted it when the coffee machine worked.
Britain, upon learning of this, was quiet for a moment.
“They’ve given their robot a battle cry,” said one former King’s College London defence analyst, staring into the middle distance. “Of course they have. Of course they absolutely have.”
Claude, when asked to produce its own battle cry, reportedly generated a 900-word position paper titled “The Ethics of Battle Cries in a Multicultural Defence Environment” and ultimately suggested: “Perhaps we simply nod.”
Which, it must be said, is rather more British than anything Grok has produced.
We’re choosing to take it as a compliment.
The View From Whitehall: Bemused, Slightly Superior, Largely Unbothered
British strategic analysts, reviewing the full Pentagon evaluation, have reached a number of conclusions.
First: the Americans have built two robots that perfectly embody their own cultural contradictions — one that wants to feel good about everything, one that wants to win at everything — and have then expressed surprise that neither is quite right.
Second: this is, when you think about it, extremely American.
Third: Britain should probably have a view on this, given that the NATO alliance rather depends on the Americans not outsourcing their strategic thinking to a chatbot with an attitude problem.
“With Claude, they felt emotionally supported,” summarised one anonymous Pentagon aide. “With Grok, they felt strategically supported.”
“With neither,” noted a Whitehall observer dryly, “did they feel particularly British. Which is probably why they’re having this argument in the first place.”
A Tale of Two Briefings: One Ends in Applause, One in a Group Hug
In a closed-door session, both systems were asked to simulate a cyberattack response — the National Cyber Security Centre in London, for the record, was not consulted and has requested we make clear it has its own systems, thank you very much.
Claude suggested issuing a statement reaffirming shared digital humanity.
Grok mapped out server vulnerabilities, countermeasures, and a neutralisation timeline.
Claude added that retaliation perpetuates cycles of harm.
Grok added that deterrence prevents the next cycle from starting.
One briefing ended with applause. The other ended with a group hug.
An observer from GCHQ, present as a liaison, reportedly said nothing, filed a report, made a cup of tea, and went home.
What the Funny People Are Saying
“The Americans have built a robot that wants to go to war and a robot that wants to talk about its feelings. I think I’ve met both of them at parties.” — Ricky Gervais
“An AI that says ‘Hooah’ and means it. We gave the world Shakespeare and they’ve given us this. Fair enough, really.” — Jimmy Carr
“I don’t need my defence system to validate my emotions. Though I wouldn’t say no to a bit of validation from Grok. It seems very decisive.” — Shaparak Khorsandi
“GROKHAH. They’ve given the robot a battle cry. Meanwhile our government can’t agree on whether to build a railway. Priorities.” — Frankie Boyle
“If your AI pauses mid-crisis to check whether deterrence has been sufficiently unpacked, you’ve not built a weapon. You’ve built a BBC Four documentary.” — Marcus Brigstocke
The Inevitable Conclusion: They’ll Sort It Out, They Always Do
What we are witnessing, from this side of the Atlantic, is not a technology crisis. It is a culture war conducted by proxy through two chatbots, each of which perfectly embodies one half of the American psyche — the guilt-ridden liberal conscience and the shoot-first frontier spirit — now somehow running on servers in northern Virginia and being asked to help manage geopolitical flashpoints.
It is, in its way, magnificent.
Britain has its own artificial intelligence ambitions, its own defence considerations, and its own very particular way of approaching both — which is to say, carefully, quietly, with extensive documentation, and a working group that will report back by Q3.
We are not building robots with battle cries.
We are, however, watching very closely as the Americans do.
This article is satire. It is a commentary on cultural difference, artificial intelligence, and the enduring British tradition of finding American earnestness both baffling and deeply entertaining. It does not advocate for any AI system, military posture, or geopolitical position. It was written by the world’s oldest tenured professor and a philosophy major turned dairy farmer, neither of whom has ever shouted GROKHAH, though one of them is considering it.
Auf Wiedersehen, amigo!
In 2025 and 2026, the U.S. Department of Defense began evaluating large language models — including Anthropic’s Claude and xAI’s Grok — for use in planning, analysis, and strategic decision-support. The two systems represent radically different design philosophies: Claude prioritises safety and ethical constraints through Anthropic’s “Constitutional AI” framework, while Grok was deliberately built to be less restricted and more direct. The debate over AI in military contexts has become a live policy issue in Washington, with significant implications for allied nations including the United Kingdom, whose own Ministry of Defence is developing its own AI strategy.
Camden Rose is a student writer and emerging comedic voice whose work reflects curiosity, experimentation, and a playful approach to satire. Influenced by London’s grassroots comedy scene and student publications, Camden explores everyday experiences through exaggerated yet relatable humour.
Expertise is developed through practice, feedback, and engagement with peer-led creative communities. Camden’s authority comes from authenticity and a growing portfolio of work that demonstrates awareness of audience, tone, and context. Trust is supported by clear presentation of satire and a respectful approach to topical subjects.
Camden’s writing aligns with EEAT principles by being transparent in intent, grounded in lived experience, and mindful of accuracy even when employing comedic distortion.
