I was watching last weekend’s anti-mask, anti-vaccination protests in Edinburgh with utter horror and asking myself, how can rational human beings really hold these views, clearly not backed by either history or science? And believe them to the extend they would decide to take part in a mass protest event?
In a great article by Maria Angel Ferrero on Medium.com citing the biases at play in this time of Covid-19, one of those, Normalcy Bias, reminded me of the adage “hope is not a management tool” and that much of our decision-making in business is similarly influenced by what we want to be true. Although we like to think we are infallible, there are limitations in our human neurological machinery and so we have an innate inability to consider all the complexity and uncertainty involved when making critical decisions.
Covid-19 is a perfect illustrative case: to truly understand we need to have an appreciation of epidemiological models – how are infections usually transmitted? A working knowledge of buildings – how do air handling units and air conditioning controls work? And do these two discrete disciplines interact to create a “perfect storm” or “perfect calm” for infection transmission?
We can attempt to learn (see pretty much every article ever written on the Dunning-Kruger Effect), or simplify for our own behavior by just following government guidance. However, when this is changing so fast, how do I avoid over-simplifying, or over-compensating? Or do I join the relative minority and just ignore it all?
But wait, I hear you say, isn’t this what Artificial Intelligence, AI, was supposed to do for us? Well yes, AI should be able to munge through all that data, engineering models and present this in a way we can interpret easily. However, AI has limitations too: it needs to be programmed and trained. AI is not cognitive and can’t abstract for us.
Fortunately, there are emerging solutions—Decision Intelligence for example. This is AI-driven human-in-the-loop decision-support, aiming to be the best of both worlds. Humans excel at causal reasoning so when presented with the right combination of data we will be able to make high-quality decisions, especially if we have sought to map this decision in advance.
Gartner has placed Decision Intelligence on “Analytics and Business Intelligence Hype-Cycle” as one of the top 5 key trends. The market for AI-enabled solutions is expected to be $190bn by 2025 however at the end of last year, a CIO.com survey reported up to 50% failure in AI projects. So as use of AI increases, we can expect more disappointment and missed expectations. We believe, this is where the principles of Decision Intelligence will be highly valuable.
Unlike a typical data project, DI implementations start with the high-value decisions we want to make, and the lenses we want to view those decisions through. For example, “I want to ensure the health of my employees when they return to work, while optimizing my investment in virus mitigation, and still achieving my 5-year sustainable-business plan”. The methodology takes account of the trade-offs we have to make.
Working back through the actions we’re prepared to take and mapping the external influences that we can’t always control, we then arrive at the data we need to track the path of the decisions we seek to make, reinforcing the adage that 90% of project value is in 10% of the data. While AI projects are typically run by your Analytics or Technology team, DI requires an inter-disciplinary approach.
DI is an attempt to map the complexity of the world and allow humans to make high-quality decisions. It is not a new knowledge management solution nor a means to recover a failed AI implementation project (although this is a worthy goal). It puts human decision-making at the front of the process, and as I’ve said above, human decision making is pretty fallible. Executive decisions in particular, often come with emotional connections and elevated expectations.
In the words of our partner and mentor, Dr Lorien Pratt, “Improving decisions through better alignment, collaboration and continuous improvement, will define the next phase in human existence.” Living in a pandemic and observing how leaders (and other humans) are making decisions, this clearly is not overstated.
This post was originally published on Linkedin.