A cartoon circulating in this space last week showed a man at his desk bragging about his AI. It converted his bullet-point e-mail into prose. He sent the paragraph to a colleague, who bragged as her own AI tool converted the text back into easily consumable bullets.

The point? Not all applications of technology are practical, or even sensible.

The cartoon’s more subtle message is particularly pertinent to insurance technology. It may be tempting to expect technology to do better than people. But automated very often isn’t superior.

In the cartoon, the hapless chap facing the daily task of writing an e-mail could have done better himself. He needed confidence in his own instincts. In insurance, technology exists to help us underwrite more efficiently, or more easily, or even more accurately. But it’s not safe to assume that AI – or any other gizmo – can do it better.

I’m an unabashed insurance technology evangelist, but deploying tech for its own sake can be wasteful and dangerous. We can enslave machines to do the repetitive and the mundane, but people and their instincts, not algorithms, deliver irreplicable value-added.