Dissatisfaction Engineering, Risk Shifting, Data Science and Business Intelligence
Are you an advice victim?
Nowadays, it feels like everybody has got advice to give. The shelves of bookstores and libraries are filled with books that promise you can achieve some goal or avoid some pain — if only you’ll follow their advice. The Internet is populated with billions of web pages, social media posts, advertisements and other message forms urging you to do something that will also help you achieve the goals or avoid the pain. Business Intelligence services proliferate to do the same at an organisational or ecosystem scale.
Of course, when I say ‘advice’ I mean any situation ranging from decision support through influencing and planning. Any situation where you may have -or wish — to rely on data produced by an ‘other’ whether that other is a person, organisation or a system.
Human beings are a social species. They have always shared data. Parents would have told their young not to eat certain items or play in particular locations because of perceived risks and observed experiences. But the Internet has changed all that. Now we receive advice — in great quantities — from people we have never met and who do not know us at all.
We can broadly divide these torrents of sage advice into two types
Advice that is based on data which has been collected in some form and treated in some way to validate it. We can call this General Personalised information (GP) because we collect lots of data examples across a population, extract a set of features to create a model and then test the data for some set of qualities (‘fit’) to make sure the model represents the individual subject.
Advice that is based on data collected through personal experience which the experience owner has generally applied consciously or unconsciously to a situation. We can call this Personal Generalised (PG) information because it takes a data source about the individual subject and generalises it across a population. Within the scientific literature — particularly in social psychology — this has a bad rep as the progenitor of stereotyping and prejudice, but it used to be the norm.
Currently the former — data driven — GP approach seems to be highly popular, because it is seen as more ‘scientific’. It purports to offer ‘insights’ which reveal new opportunities and -sometimes- hints regarding the tactics that can be used to realise them. We imagine that it acts as a scrubber to remove the so called- “biases” that haunt human decision making, leaving them all clean, shiny and effective- decontextualised in fact.
And the latter — experience driven- PG approach seems to be fading somewhat as increased transparency shows the costs of poor personal decision making and the need to manage risk by sharing blame, preferably before everything goes wrong.
We used to be able to balance GP and PG information. Now, sadly, it isn’t always that easy. Decision makers who don’t use GP decision making have to look over their shoulders. Decision makers who do, have new stakeholders to consider, and an ever-growing toolbox with -sometimes -unfamiliar tools.
So, what do we do with this clean, shiny, effective new data, which may, after all, be untouched by human mind as it passes through the clean rooms of Data Science, Artificial, and Machine Intelligence and Deep Learning?
Well, in some cases, we can feed it directly into Supply Chain or ERP systems, where the integrity of the information is preserved right up to shipping.
But mostly, we give this data to humans where it immediately becomes data (advice) that is ‘contaminated’ by human interpretation and use. Issues such as politics, self-interest, greed, stupidity, and all the other interesting human characteristics come back into play, and after all the effort of cleaning, data dirt once again fills your environment.
So, is it worth it?
Will it solve your problems?
Or is data (advice) just another distraction meant to increase your dissatisfaction with your environment in the sort of FOMO exercise that manufacturers of cleaning, decorating and furnishing products encourage us to use in the home?
And can your house of data be so ‘clean’ that it has lost the only thing that made it liveable?