Deviens ce que tu es. Become who you are. That was the advice of Lacoste about 30 years ago, lifted directly from Nietzsche, master of the self. For sure when I wore my first overpriced polo shirt with its little alligator on the breast, it made me feel different, someone apart (bet in retrospect not in the least more me).

I saw more and more people had become who they are. They were being their unique selves and expressing their individuality by displaying congregations of the embroidered reptiles wherever you looked. Then I had a revelation: a good thing that sets you apart can very quickly become simply commonplace.

The head of a big reinsurance group said recently on his company’s half year earnings call that the financial benefits of AI would eventually follow this pattern. Everyone would have it, so it would deliver no particular advantage for anyone. AI would become table stakes, like every other technological advance since the typewriter.

I’m pretty certain that’s not the case.

First of all, in the rush to embrace new technologies, it’s easy to overlook the fact that they are simply tools. Those with better tools should, all things being equal, always be able to perform better than those using sub-standard equipment (although it isn’t always so; try passing through a door with a keyless entry system when the power’s out and the battery’s flat).

In principle, eventually everyone can have the best tool for the same price, which in theory should eliminate any potential advantage that can be gained through technology. That was certainly true of the typewriter, and of its great grandson, Microsoft Word. But it isn’t necessarily true of technology for insurance.

The second point is key: technology is simply a tool. Data is what gives underwriting shops an advantage. Those that have the best data, which means the most appropriate data, the cleanest data, the fastest data, the most accurate data, and the data that’s most accessible and relevant to the task at hand, are the ones that will enjoy the greatest success, provided they’re tools for underwriting and – critically – the human skills and experience that oversee them are market-leading.

In our world, the world of specialty and wholesale insurance and reinsurance, lowest-common denominators don’t cut the mustard. Everyone’s data must be exactly the right data for the peculiar risk before them, whether it’s the correct satellite images, the specific telltales from 20 years’ worth of 10-Ks, or some other nugget drawn from an ocean of bits and bytes.

One-size-fits-all technology cannot deliver that intelligence.

To succeed, underwriting outfits need to implement the architecture and platform that allows them to record as much relevant, curated data as possible, to drive superior, shop-focussed analytics that support the unique underwriting unit’s risk appetite and approach.  The more data they have, built up over years, the better they will perform.

That’s accomplished through relentlessly improving pricing and risk selection, again fuelled by technological solutions which fit the challenge and the users, not the vendor. Meanwhile, underwriting teams must ensure that they automate as many of the non-underwriting tasks as possible. The goal isn’t simply or even primarily cost reduction (although that’s a great side gain). Instead the key benefits are improved data quality, the ability to consume and parse larger volumes of data, and – the human benefit – to attract high-quality underwriters who love the idea of spending most of their time on underwriting, rather than on manual repetitive tasks.

There’s another piece to the puzzle. You can have the best insurance technology tools in the world, but if you don’t know how to use them, or perhaps more accurately if you don’t use them well, you may as well have a stone axe. I refer back to Real Madrid and the Galácticos, whom I’ve written about here before in this space.

Assemble the best footballers in the world, the plan was, and they could only win. Unfortunately the superstars’ individual brilliance isn’t the only ingredient for success in a team sport. It cannot compensate for the absence of system, structure, and clearly defined roles for everyone on the team. Club president Florentino Pérez Rodríguez Had the best tools, but he wasn’t using them well enough.

Similarly, the most powerful insurance technology tools available will not, on their own, lead to a revolution of results. That requires those tools to be used well, to resolve genuine pain points, and to provide the foundation for genuinely new, more efficient processes to be created (rather than using them simply to patch up old approaches that haven’t really worked very well so far this century).

AI is one of the tools that can makes all this possible, but for many applications, other simpler tools will work even better, alongside or instead of an AI agent. Sometimes all you need to open a door is the key in your pocket. Underwriting organisations which give people such keys will surely flourish. They’re the places where the stars of our sector will become who they truly are.

* Like every Insurance Technology Diary entry about AI, this one is accurate only to the best of my knowledge at the time of writing. The pace of AI progress is so great that I cannot guarantee it remains so now that it’s finished, let alone when you read it.

Guillaume Bonnissent is CEO of Quotech.