Brand value and conceptual novelty have almost nothing to do with each other. That's the finding. Here's what produced it.
I built a project called Modern Retro. 100 modern brands, reimagined as 1976 high street shops. Spotify as a record store. Coinbase as a bureau de change. OpenAI as a research institute. It started as a visual experiment -- what do these brands look like through a Wes Anderson lens, shot on 35mm Kodachrome, warm tungsten light? Then came the films: each shop rendered as a short Super 8 reel, grain and hiss, a brand that never existed in 1976 briefly made real. But somewhere in the process of building both, I needed a way to rank how absurd each reimagining actually was. So I created the MRA Score: Modern Retro Absurdity. Three dimensions -- Era Dissonance, Cultural Distance, Concept Delight -- each scored out of 10, averaged to a single number.
The scores made the finding concrete. And the finding is stranger than it sounds.
Nike scores 3.0. OpenAI scores 9.5. The distance between them isn't money or culture. It's vocabulary.
Nike scores 3.0 on the MRA. One of the most valuable brands in the world -- and one of the least absurd as a 1976 shop. Because the idea is ancient. Make shoes. Sell them to athletes. Make those athletes feel like winners. You could explain what Nike does to a shopkeeper in 1976 in about thirty seconds.
OpenAI scores 9.5. Try explaining what OpenAI does to that same shopkeeper. Not the technology -- just the idea. A machine that has read everything humans have ever written, and writes back. A machine that reasons. A machine that codes. Before you get to the product, you need to explain computers, then software, then the internet, then neural networks, then what "training on data" means, then why that produces something that can hold a conversation. You'd lose them before the first sentence landed.
That gap -- between what a brand does and what a 1976 mind could comprehend -- is what the MRA Score actually measures.
The score isn't about disruption. It's about whether the idea existed before.
Disruption is the word we reach for when a new company does an old thing better or cheaper or faster. Uber didn't invent taxis. It disrupted them. The MRA for Uber is 7.3 -- notable, but not at the ceiling. Because you can describe Uber in 1976: it's a taxi company, but you book it instantly and a stranger who owns their own car drives you. That stranger not being an employee is the new bit. One new concept inside an old frame.
Compare that to ElevenLabs. Voice cloning as a service. The studio has no singers. The voice exists as a digital pattern, applied to any text, in any language, at any time, for a fraction of the cost of hiring a human being. There is no 1976 frame for this. The closest analogy -- a recording studio -- is accurate about the output and completely wrong about the process. ElevenLabs scores 9.5.
The difference between 7.3 and 9.5 isn't market capitalisation or cultural impact. It's whether you need new words to describe what the company actually does.
Crypto scores lower than AI. This is the most contested call in the ranking, and I think it's the right one.
The instinct is to push back: crypto is new. Bitcoin launched in 2009. ChatGPT launched in 2022. Surely recency maps to absurdity?
Not quite. What matters isn't the launch date. It's whether the underlying concept has a 1976 equivalent.
"Exchange foreign currency" is a concept with a 1976 analogue. Bureau de change existed on every high street. The currency Coinbase exchanges doesn't physically exist -- it lives on a distributed ledger maintained by thousands of computers simultaneously -- which is genuinely strange. But the function, convert one unit of value into another, is ancient. A 1976 shopkeeper told about Bitcoin would be confused. But the shape of the confusion is manageable. They have a frame for the category, even if the specific instrument baffles them.
Now try explaining a large language model. The confusion isn't about the mechanism. It's that the category doesn't exist yet. There is no frame. You'd be building vocabulary from scratch before the product description could even begin.
That's why AI scores above crypto. Not because it's more disruptive. Because the idea itself is more new. I've had this argument with people who know crypto better than I do. I haven't changed my view.
The brands at the bottom are not failures. They're just very good at old ideas.
LEGO scores 3.0. The company has been making construction toys since 1932. The bricks, the concept, the delight -- all of it would be entirely at home in a 1976 toyshop. Visiting LEGO in Modern Retro feels less like satire and more like a portrait.
Patagonia scores 3.0. Founded in 1973, three years before the project's canonical year. Outdoor clothing for serious climbers. The ethical supply chain and the B Corp certification arrived later, but the core identity -- quality kit for people who spend time outside -- is period-accurate. A Patagonia shop in 1976 is not a reimagining. It's almost a photograph.
Aimé Leon Dore scores 3.0. A vintage menswear boutique reimagined as a vintage menswear boutique. The score rounds to zero.
None of these are weak brands. They found genuinely good ideas and executed them with discipline across decades. The idea doesn't need to be new. The execution just needs to be right. The MRA Score doesn't punish longevity -- it just can't reward it.
What the thought experiment reveals about right now
We conflate a lot of things under the word "innovation." New product. New model. New market. New technology. New behaviour. What we rarely stop to ask is: does this require a new idea, or a better version of an old one?
Most successful companies are the latter. Better distribution. Better design. Better economics. A genuinely new idea -- one that requires new vocabulary before it can be explained -- is rarer than the noise around it suggests.
In 2026, the list of ideas that require new vocabulary is short. AI is at the top of it. Not because of the investment cycles or the headlines. Because you cannot describe what a large language model does to a 1976 shopkeeper. Not in thirty seconds. Not in thirty minutes. The concept itself arrived recently enough that we are still building the language to contain it.
The MRA Score started as a piece of fun attached to an image project. A way to rank how absurd each imagined shop was. It ended up answering a question worth asking: what is actually new right now, in a way that has no precedent?
The answer, if you're honest about it, is one word. AI.