We all know the story of Orson Welles' infamous radio broadcast of The War of the Worlds.

On Halloween Eve in 1938, Welles dramatised H.G. Wells' novel about an alien invasion so convincingly that listeners genuinely believed the world was being attacked. The result was widespread panic across the US, with people taking to the streets in fear.

It's the most famous radio hoax ever, and from the 21st century it can seem pretty hard to believe. And yet we believe it because that's what the newspapers of the time said.

Except, here's the thing: the panic never actually happened.

According to Derek Gibson, in his recently released book Data, AI and Noise: "The story of the hoax is a hoax itself, driven by misinformation."

By digging into the data, Derek proves it couldn't possibly be true. Which raises the question: if an "accepted fact" like this is actually untrue, how many other "accepted facts" are also false? And how many of our stakeholders' beliefs are equally untrue?

A data scientist I know recently told me a story where this was exactly the case. His stakeholders were convinced that an intervention they'd introduced was working, and they were continuing to invest in it as a result. But nobody had bothered to check.

So he quietly went away and did just that, and found no evidence that it was having any effect. When he told them, they were shocked, but the data was clear. He saved his stakeholders lots of money, and he became the hero of the day.

AI is making it progressively harder to separate fact from fiction, and this is where data scientists can help. Not by building better models, but by asking what "facts" stakeholders rely on, and then looking at the data behind them.

So next time someone says "we all know...", don't take it as given.

As the saying goes: "In God we trust. All others must bring data."

Happy Halloween and talk again soon,

Dr Genevieve Hayes