Radical Uncertainty

Originally posted on July 19, 2020 @ 9:32 AM

Radical Uncertainty

imageI’m not an economist but we have much we can learn from other disciplines. Kay and King’s book Radical Uncertainty (2020) should be a compulsory read for safety people. Moreso, for anyone spruiking the nonsense of zero.

Kay and King have produced an easy to read book on the nature of uncertainty, the nature of human cognition and the way humans make decisions. They show how human thinking is composed on narratives and how some narratives do not serve us well in tackling risk.

They start the book with some big picture surprises, failures and mysteries with which we are familiar, like: the decision to raid Osama Bin Laden’s lair, Churchill’s decision styles and the Steve Job’s Second Coming. They do so to draw a difference between solving puzzles and tackling mysteries and make the point that the framing of both is critical for success. They show through the stories of Wang, Nokia, Blackberry, Kodak, IBM, Microsoft, Lehman Bros etc. just how volatile and uncertain the world is. They draw on that famous speech by Donald Rumsfeld who cited the ‘unknown unknowns’ to justify the attack on Iraq and pull apart the constructs and narratives humans compose to assert certainty, when there is none. They show how false faith is placed in ‘probablistic reasoning’ and the delusions of mathematical certainties.

Several chapters discuss the assumptions and weaknesses of ‘behavioural economics’ and particularly the assumptions of Kahneman and Tversky. The assertion that human biases and heuristics are a deficit or flaw in human disposition is pulled apart most effectively. They also look at the work of Klein and other populist models like ‘nudge’ theory and the vacuous nonsense of vision and mission statements. From my perspective it seems that we tend to name such documents by what they are not. I’ve rarely seen a vision statement that is visionary. The more conservative and compliance –focused an industry, the less vision there can be.

Kay and King are not anthropologists but spend a number of chapters trying to understand the worldview of anthropology, including some experiments where anthropologists and economists are given the same problem. They also look at legal narrative reasoning and again compare this to how economists construct understandings of risk. The outcome of their studies demonstrate how different worldviews and transdisciplinarity enhance understanding and enlarge strategies for tackling risk.

‘The spirit of enterprise dies when mathematical expectation takes over’ is a neat quote that frames the way Kay and King explore the limits of numerics and metrics in understanding risk. Considering the date of the writing of this book and publication they state this:

‘But we must expect to be hit by an epidemic of an infectious disease resulting from a virus which does not yet exist’. (p.40) as a way of framing a discussion about the false certitudes of ‘behavioural economics’. The discussion shows just how stupid it is to talk about predictive analytics, big data and ‘future proofing’ against the realities of ‘radical uncertainty’.

The book is structured in five parts and the final part is focused entirely on economics and risk. There is some helpful material in section 5 but the best stuff for me was in the first four parts of the book.

It was good to read Kay and King pull apart the nonsense assumptions of scientific method and the crazy assumption that validity is limited by repetition and quoting Max Planck who stated ‘science makes progress funeral by funeral’. This section on ‘false stories and bogus statistics’ reminded me so much of the concocted attributions of the safety industry and its religious-like belief in bogus models that don’t work.

So, if you are interested in the realities of risk, the nature of human decision making under uncertainty and, pulling apart some of the sacred cows in thinking about risk, this is worth reading.

Source link

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.