Data science, high performance computing (HPC) and machine learning algorithms allow organizations to make better decisions and create game-changing strategies. The integration of high quality smart data, computing resources and data scientists using learning algorithms is the secret sauce to achieving the fundamental goal of creating durable competitive advantage.
Organizations with the best data scientists and learning algorithms (AI) have a significant competitive advantage. Professional data scientists use artificial narrow intelligence for massive advantage. Machine learning algorithms can add great value in certain circumstances within appropriate contexts and within data science processes.
Seasoned data scientists learn to love uncertainty. Making high quality decisions under uncertainty creates massive competitive advantage. The more limited information and uncertainty the greater the advantage for the professional data scientist.
Acquiring awesome data and applying probability theory is optimal strategy when making decisions under uncertainty. Over time, high quality decisions will provide consistent probability advantages and mitigating risks (reducing bad decisions) will lower probability of blowing up.
The nature of war has changed. Traditional kinetic warfare is still operational yet evolving from massive air carriers and big tanks with large ground forces to smaller special forces relying on near real-time intelligence to achieve specific goals within macro strategy.
The easiest person in the world to fool is yourself. Data scientists sometimes fool themselves - in matters trivial and important. Thus, I suggest we acknowledge real or subconscious biases in ourselves, the data, the analysis and group think. It is prudent for data science teams to have both internal and external checks and balances to expose potential biases and better understand objective reality. Here are a few ways data scientists sometimes fool themselves:
Confirmation bias: tendency to favor data that confirms beliefs or hypotheses.
The Nuremberg Code was drafted to protect humans from medical experiments based on exploitation of prisoners by German scientists and physicians during World War II. Ten (10) elements of Code:
Below are links to the codebase of the major global COVID19 forecasting models to compare predictions and evaluate predictive performance:
Modern day concepts of "intelligence" and "meritocracy" appear to be based on a single test taking metric. Those who excel at test taking gain entrance into purported "elite" institutions and thus have a "moral" and "merit" based mandate to lead. Yet strong reality based evidence suggests those with superior test taking ability are NOT most fit to command public and private organizations. Real world experience casts doubt on current notions of "meritocracy" and the modern day leadership class has lost the mandate of the heavens.
Economists apply theory to complex reality yet fail to accurately interpret the past attempting to forecast the future.
Lawyers define and manipulate words yet fail to define justice within legal architectures.
Physicians apply medical science to cure disease yet fail to define health and accept unknown unknowns.
Accountants count yet fail to calculate future uncertainty.
Mathematicians apply logic yet fail to factor human fallibility.
Psychologists apply academic theory to complex human behavior yet fail to understand their own minds.
Seasoned data scientists avoid the quantitative fallacy trap where you focus solely on certain quantitative metrics while ignoring other non-quantifiable variables. While the old saw that you cannot improve and manage what you cannot measure is true - what you decide to measure and not measure matters a great deal for understanding complex static, situational and fluid reality.