Canary Wharf is not where I expected to have my own understandings of risk management challenged but life is strange like that. In a meeting room in the KPMG office, I saw the value of quantitative risk for the first time. It also loosely coincided with a realisation that artificial intelligence was something that, if used correctly, could be an incredible tool.

And here I am, a convert to both quantitative risk and artificial intelligence.

I’ll start with the less controversial one of the two: Quantitative Risk. For most of my career I’ve used impacts and likelihoods to place risks on five-by-five matrices to understand whether a risk is High, Medium, or Low. Wherever possible I made sure to make the process as objective as possible.

But the cracks were there. Describing the right-hand corner as the roof is on fire and you are pouring on petrol box should have been a clue. The lack of granularity. The lost sleep caused by being told to make an asymmetric scale work.

Quantitative risk management is different because it’s no longer about agreeing what High or Almost Certain means. You determine the impact from real world data and probability replaces abstract likelihoods. Instead of conversations trying to explain what it means to be Low instead of Medium, you show the specific changes. Perfect? No. 100% accurate? Not a chance. But better.

And so, to artificial intelligence. To misquote Homer, it promises to be both the cause of and solution to all of life’s problems and this whole site is testament to my belief in the values of AI, so I won’t go on. Perhaps it’s because I am lucky enough to work with some of the smartest people in the field, who have shown me the promise of AI that I really do believe that this a tool to be harnessed.

Hell, this site likely would not exist if it were not for AI. I’ll let others judge whether that is a good thing or not.

David