Now that AI has made polished design and perfect prose the norm - flawless no longer equates to quality - it simply means you have the same subscription as everyone else. When flawless is everywhere, authenticity becomes the rare commodity: the judgment, values, and thinking layered on top, none of which is available on subscription.
Last week, in a meeting room at Cape HQ Alex was running us through a presentation, gathering feedback. Clear, well-structured, nothing obviously wrong with it. And then, almost without registering it, I noticed the punctuation. An em dash. Then another. Large language models use em dashes (the long punctuation mark used to replace colons and other sentence breaks, named as such due to its typographic length matching the letter m as opposed to the shorter, en dash) - the same way anxious people use filler words: compulsively, even when asked not to. The first thing any student does before submitting an AI-assisted essay is run a 'find-and-replace' on them. It is the digital equivalent of removing the price tag.
When I made my point, Alex pushed back immediately. He had been using em dashes his whole career. Well before AI made them fashionable, and then subsequently deeply unfashionable. He was right. He had not changed. The world had simply caught up to him.
AI design tools did not invent the clean sans-serif, the generous white space, the italic emphasis words, the restrained palette. It was trained on thousands of premium brands, ours presumably amongst them. So, when a Cape deck and an AI-generated deck start to resemble one another, the honest reading is not that we have been swallowed by the machine. It is a process that repeats itself time and time again:
1. The soulless grid
Imagine a Roman legionary, reassigned for the third time in five years, arriving somewhere he has never been. Lugdunum (today Lyon), say, or Londinium (London), or a garrison town on the edge of nowhere with a name nobody outside the empire could pronounce. He does not need a map. He walks in, finds the cardo maximus running north to south, the decumanus crossing it east to west, and within minutes he has located the forum, the baths, the market, and, more importantly, the nearest place to get a decent meal. Every Roman city ran on the same two axes because Rome had already spent centuries working out, through trial, fire, and the occasional catastrophic drainage failure, what a functioning city required. It exported the answer. Rome invented the city grid. It was not laziness. It was the distillation of everything that had worked, optimised and reproduced at continental scale. Tothis day, we admire the Roman cities, the roads, the infrastructure but it came at a cost; the provincial cities are soulless, each coliseum, each theatre, is the same. We visit them all but fail to distinguish one for another.
Claude Design, or any AI-driven best-practice synthesis, is the equivalent of the Roman engineering blueprint. It surveyed ten million presentations, ten million pitch decks, ten million investor memos, identified the axes that worked, and handed back to everyone the same urban plan. Beige background. Burgundy italic header. Serif body, one accent font. The forum is always in the same place. You can find the key message without asking for directions.
2. Ninety-five arguments that changed Europe
Gutenberg's story usually stops at "the printing press changed everything," but the better part comes later. Fifty years after Gutenberg, printed books across Europe looked almost identical: the same font, the same layout, the same religious pictures on the cover. Put a hymnal from Basel next to one from Antwerp and you would barely tell them apart. The look had become standard. Almost boring.
Then, using that same standard printing press, Martin Luther printed his ninety-five arguments and pinned them to a church door in Wittenberg. The paper looked like everyone else's paper. The font looked like everyone else's font. Nobody holding that pamphlet would have guessed, just by looking at it, that it was about to split the Church in half and reshape Europe. The press did not create Luther's ideas. It simply carried them, printed the same size and shape as everything else on the shelf. The format was the same for everyone. The thinking was not.
3. Times New “Roman”
In 1929, a typographer named Stanley Morison publicly complained that The Times of London looked old-fashioned and badly printed. Instead of ignoring him, the paper hired him to fix it. Morison built a new typeface for maximum legibility and economy of space, and unveiled it in 1932 under the name Times New Roman.
The "Roman" refers to an old habit: Renaissance printers, five hundred years earlier, had modelled their clean, upright letters on the inscriptions carved into Roman stone monuments, the same instinct for order and legibility that laid out the cardo and the decumanus. Rome's real export was never a single city plan. It was a taste for clarity that later generations kept reaching for, in stone, in print, in font. Times New Roman went on to become the most reproduced typeface in history, sitting quietly behind millions of essays, contracts, and government forms.
But, the oldest test still holds
Our brand lead Tess recently reminded us, that a brand, like a person, is not its clothing. The fonts and colour palettes are the shop window; and now anyone with the right subscription can build a passable 'brand design' in an afternoon. What cannot be copied so easily is what happens inside the shop - the true brand identity: the values, the culture, the judgment, the delivery, the way a person in the room reads a client's hesitation and adjusts before anyone says a word. This is also why Tess is right that 'perfection' in design and editorial text is starting to look less like quality and more like a warning sign. When flawless is cheap and available to anyone, flawless stops reassuring and starts raising questions. People now start to look for evidence of authenticity, so the deck that looks a little too polished triggers the obvious red flag: is this the person's own thinking, or the interface talking? The oldest test still works, because it has never needed updating: if you cannot sketch the argument on a napkin, with a pen, and still get the attention of the audience, the slides are doing the work you should be doing yourself.
None of this argues against the tools
Cape uses AI extensively, and should. The research runs deeper for it, the data points sharper, the drafts faster. But there is a difference between using a tool well and redesigning the brand every time its aesthetic shifts. That race has no finish line, because by the time you have caught up to this season's AI look, the AI will have already moved, having learned from whatever the best brands did next. Chasing it is not differentiation. It is corrosion, one small capitulation at a time, and Tess has been the one insisting we do not enter that spiral.
Breaking the glass ceiling
Here is the paradox: the better the tools get, the harder it becomes to look different, precisely because everyone now has access to the same tools. AI raises the floor for the whole industry. It does not raise anyone's ceiling. You have not got worse. Everyone else has simply caught up to where you already were, which, seen from the client's chair, looks exactly the same.
Rome built the grid because the grid worked everywhere, for everyone. What it could not export, along with the roads and the drainage, was Rome itself; the ambition and appetite that made the city more than its infrastructure. Luther's pamphlet came off a press that printed a thousand forgettable others that same year, on the same paper, in the same font. Nobody has read those. Times New Roman sits behind more documents than any typeface in history, and not one of them is remembered for its font.
Cape can use every advantage AI offers and still be recognisably, stubbornly itself, for exactly the reason the Roman provinces never became Rome no matter how faithfully they copied its streets. The grid was always available. What built the city was everything laid on top of it, over years, by people who had opinions about what belonged there and were willing to defend them.
The ceiling was never made of glass. It was made of judgment, and judgment does not ship with the subscription.
William Raynar