Quite a bit was really cool about the Ptolemaic Dynasty. They were the royal family that ruled the Kingdom of Alexandria in Egypt for about 275 years, right up until Elizabeth Taylor, Queen of all Egyptians, fell in love with Richard Burton, and handed it all to the Romans.

I saw the coolest Ptolemaic thing first hand during my first summer-vacation visit to the British Museum, with all its Tutankhamen tatt. It’s the big black hunk of rock called the Rosetta Stone, which unlocked all the secrets of the entire Egyptian civilisation.

The Stone was rediscovered in 1799 by Pierre-François Bouchard, a polymath and countryman of mine, who was visiting Egypt that year on a sort of cultural exchange. He immediately saw the rock’s potential. It has, at the top, several lines of hieroglyphs, which to that point had been all Greek to scholars, who could not decipher them for toffee. Next came several lines of Demotic, a more exclusive form of Egyptian former writing which has no connection to the Devil despite its name. Finally came many lines of Greek, which despite being all Greek, were widely understandable.

The thing that made the Rosetta Stone so valuable is that all the bits of text said the same thing. It was only a boring statute, but it allowed Jean-François Champollion, who (as I had learned, like every other French schoolchild has learned) managed in 1822 to decipher the stone’s ancient Egyptian, which everyone had forgotten. Now, relatively suddenly, we could translate back and forth between ancient Egyptian and English. Translation tables, the precursors of Google Translate, made it almost effortless.

Now, of course, we have Google Translate. Everyone has tried it, if only to look impressive when booking a table at a Portuguese restaurant. The software behind it is really cool: as long as the language is known, Google can translate it. It can learn new languages easily, too. Give Google a relatively consistent new language, along with a body of writing and a ‘Rosetta Stone’ in a language it already knows, and the new language can be cracked in a few hours. That’s really cool.

There’s no need for a ‘pivot language’ like the European Commission used to use, translating almost everything into English, French, or German before translation into one or all of the languages of the other member states of the EU. That approach used to make sense, but it doesn’t now. Google doesn’t have to translate everything into Universal. No one does, in fact. You can go direct from Japanese to Swahili.

All of this makes me wonder why people in the London market are still so excited about data standards. There’s a lot of central pressure to adopt Acord, one of the many data standards that has grown popular in the insurance sector. Unfortunately that carries a significant cost, even just to see what those hallowed standards are, and ignores a simple fact: Acord standards could be wrong for your business.

It is, of course, critical to have data standards within your own organisation. They make data analysis very much more straightforward, and dramatically enhance the possible understanding of the data you have. If your data isn’t standardised, launch a project now. If you investigated standardisation some time ago, and found the process to be cumbersome and painfully expensive, and likely to yield only mediocre results, look again. Now, with the advent of artificial intelligence conversion tools, it’s neither particularly difficult nor expensive, and results are exceptional. The tech has come a very long way, just like Google Translate.

But you needn’t adopt a set of data standards that don’t suit your particular class of risk, your business model, or your processes. To do so would be daft. You must choose (or devise) a standard that works perfectly for you. In the way that every business sector has its own jargon to describe that industry’s unique variables, functions, and processes, each type of underwriting, and each organisation that underwriters needs its own clear standards to define the data it uses.

When it’s necessary to translate data in order to share it with others, it’s not necessary to convert first to Universal. No pivot language is needed. Artificial intelligence can translate almost any data format into almost any other, and vice-versa, with just a Rosetta Stone to show it how. It can do so in real time, on the fly, as part of the data exchange.

Real-time translation devices that use artificial intelligence and voice recognition to translate spoken languages instantly already exist. They’re Uhura’s dream device, but we don’t have to wait for the 23^rd century. You can buy them on Amazon right now. Basically the same process works on your data.

At a recent market pep-rally event for Acord standards, one speaker said: “If companies had been doing this [digitising] in silos, it wouldn’t have worked. If everyone had gone off and developed their own standards, it would have fallen flat.”

That was perhaps once true, but not now. Data standards make processing more efficient, and improve the accuracy and understanding of data, and therefore of underlying risk. They’re a no-brainer. But they don’t need to be a lowest-common-denominator proposition that isn’t the best for every organisation. We don’t need Universal. That’s yesterday’s proposal. The present is already very much cooler.