Artificial Intelligence & Machine Learning: Data is King

 

I am far from being an expert in artificial intelligence (AI) and machine learning (ML). Actually, I spent some time googling these two concepts which - the truth to be said - are commonly used interchangeably. The ultimate goal of AI is to create intelligent machines that simulate the human thinking capability and behavior. Deep Blue, the famous supercomputer that defeated chess world champion Garry Kasparov falls into this category. On the other hand, ML is the art of these machines to learn in real time from all gathered data without being programmed explicitly. Most subject matter experts state that Deep Blue cannot be considered an example of ML because it was programmed to beat humans but learned little along the way.

ML has evolved tremendously since the Deep Blue times in late 1990s. Its skyrocketing advancement poses novel challenges in our lives, especially when it comes to trust. An example of this - perhaps not the most relevant, yet illustrative - can be observed in Nascar races. In them, AI and ML play a vital helping hand in understanding a massive data set, such as identifying anomalies and contributing causes in real-time. The algorithms analyze the real time data and yield the best course of action to win the race: optimum timings to tank or change tires, best time to overtake a rival, etc. In one of the races, the machine advised to do A, yet the team went with their gut feeling (they knew better!) and picked B. They lost and realized that option A, indeed, would have been a far better choice.

At the end of the day, AI and ML require above all just one thing, data. And a project generates a massive amount of it. Having in mind the DIKW pyramid, data is treated to obtain information, which is then further processed into knowledge and finally wisdom. How this translates to project management? One can think of a situation that project managers often come across during a project: making scenarios. The PM is responsible for gathering and process all relevant inputs from SMEs or any other suitable sources and present the various options with their cons and pros to the sponsor, steering committee... It is easy to envision a machine (or software) that is able to not only analyze the data and define possible scenarios but also to provide timely alerts to avoid certain less favorable scenarios, thereby increasing the odds of delivering a successful project. This specific case example wants to reflect on the adapting role that the PM must face as the AI/ML technology becomes more mature.

Cab drivers will become obsolete when self-driving cars become available at mass scale level. The threats that AI/ML will exert in project management is yet to be seen. Can they live in perfect harmony? Have your saying in the comments section below.