Presentation what if the whole world is bad in data-driven decision-making - final

Editor's Notes

• #5: http://www.oecd.org/economic-outlook/june-2020/ https://data.worldbank.org/indicator/NY.GDP.MKTP.CD?view=chart Data-based lockdowns cost 5%-10% of global GDP!– 5-8 trillion dollars 4-6 times Scandinavia GDP vanished! Over third or even half of euro-area! About same or much more than latin america and caribbean combined!
• #22: Understand and frame the problem context Prioritize development to separate meaningful investments from meaningless waste of efforts Focus on reality of measurement for relevant questions Understand the contextual validity of data collection and decision-making Productize the information production process
• #23: Data entry errors and statistics errors get attention but can be minimal when considering the trends unless the information production process is truly dysfunctional. It’s more important to understand trends, such as exponential vs linear, or relations between variables than their individual values. Data semantics do not get enough attention but are much more severe problem. Everyone knows these exist but it’s more complicated to fix these perspective differences than actual organizational errors. Died to covid-19, died with covid-19 diagnosis or died to something else. These are easily manipulated for secondary use cases. Measurement errors are massive problem as they might be completely outside of powers of individual organizations or even skills of human societies. We simply do not know or measure such things. Unknown measure levels due to lack of instruments. Inherent error rates and biases in measurement methods. The lack of systemic and contextual understanding is by far the most important problem. Partial understanding of system dynamics and hidden biases are becoming a massive societal risk as digitalization transforms everything to measurable subsystems. Unknown variables in a wider context. Biased priorization in a wider context.
• #25: If we are bad in it, we would be wasting massive amount of wealth, health and happiness for generations. Unfortunately, we do not know yet how much we have lost due to our problems in data-driven measurements, analytics and decision-making or how much we have saved lives and wealth by being enough good to avoid total catastrophy. We know that the world was not prepared for this kind of data-driven decision-making and even less for reactions that it would need societies to implement into their everyday lives. Luckily, we have a lot of real experts who are studying antibodies, total death rates and similar less public variables. It’s the public media and discussions that focus on one-sided perspective of diagnosis and diagnosed deaths. It’s the governments that manipulate the data and efforts to maintain political status quo. The true problems are in framing the problem and getting all the variables. They are in manipulating measurement and semantics. Not really in the changing statistics or data entry errors.