Leadership is about securing followers and more importantly maximizing the collaborative strength and social ties of others. Never has the power of a tribe or collective intelligence been more important than in the Age of Artificial Intelligence and now increasingly the Age of The Home.
When I was a partner at Accenture, I worked with incredible leaders like Dr. Bill Ives, a brilliant Harvard Graduate, now retired, who understood the power of social networks enabling knowledge flows in simulation structures, and Rick Stuckey, Global Partner Head of the Knowledge Management and Collaboration Centers of Excellence who understood the power of technology, process integration and the power of user centric design. We were fortunate at the time to also have Dr. Thomas Davenport, renowned author, and professor at Babson College, who led the Accenture Thought Leadership Center where many of us collectively advanced Knowledge Management and Collaboration Commerce approaches to build more trusted collective intelligence awareness. A great deal of this pioneering thought leadership can be found in the research of bees and understanding swarming as a collective organism.
What is interesting to note is that researchers from the University of Sheffield discovered that honeybees in a colony work the same way that neurons in the brain work to better understand human behavior. Their pioneering research demonstrated how honeybees can decide where to build their home or nest and also how the bee colony works as a super organism which displays a coordinated response – similar to the human brain. In other words, how bees speak to each other and make informed decisions is comparable to the way our individual neurons in our brains interacts. Prior research has also validated that single brain neurons do not obey laws, but the whole brain does.
I have always been intrigued by the power of social networks , reciprocity and trust are deeply embedded in network nodes (perhaps in neurons) and this is where I anchored my early doctoral research where I learned about the underlying power and dynamics of complexity science which drive the ebbs and flows of everything. Then somehow I ended up over twenty years writing or co-authoring 13 books on collaboration, social media, eCommerce, and knowledge portals. I recall Dr. Kumar Murty a friend, and currently a Board Director with the renowned Field Institute of Mathematics, and Founder of the PerfectCloud.io who said to me : “Cindy everything in life can be distilled into a mathematical equation.”
Currently, my research is tied into the applied leadership of designing, building and implementing Artificial Intelligence Systems, using a variety of AI: predictive and prescriptive analytical methods. The underlying motivation or my research and writings currently are to help advance the knowledge and practices of Board Directors, CEOs, and C Levels to get their data practices lined up more rapidly, so their AI models are sustained versus are not just beached and isolated models.
Almost very organization I engage in, I consistently find major gaps and risks in their data assembly and production practices. As a result, I have started to hypothesize that one of the primary reasons for these consistent big data gaps is simply that many current Board Directors and CEO’s are not asking the right questions of their leadership teams and are also not skilled sufficiently in these areas to lead in these newer areas.
But when you think about this more, few of the current Fortune 500 leaders have been schooled in integrated business, and advanced AI and Big Data practices.
How many of our top board directors and CEOs governing our F500 companies have personally led the development of an AI system – from top to bottom and sustained its evolution over the past five years? Ask this question and this may help provide your organization with new insights and learning pathways.
Understanding how adults learn is key to closing this gap.
First, how do adults learn?
Let’s not forget that adults learn by doing, by application and experiences, and they interpret skills, ideas and knowledge through the medium of life experiences and then test them in real life settings.
So delegating to others all AI projects, under the assumption that the C suite and board directors will be able to lead data transformation evolutions using AI methods efficiently may well explain the high failure rates of large scale AI projects, currently estimated at over 75% failure rates, due to lack of last mile sustainability investments.
So why not think outside the box, and assign C Levels to be immersed into the details of their AI investment programs, not just from a check in perspective like traditional sponsorship reviews on outcome and issue/risk status. Rather think outside the box. How about getting them into being immersed in full life cycle programs to learn from the design, development and evolution of their AI models.
How do we expect older generations to get data practices right, and complex AI solutions right if they are not put into deeper learning situations?
I have asked multiple CIOs recently what data management practices are they using? I never seem to get a clear answer. I mention DMBOK, and I often get a confused look. I also often speak to data management transformation leaders and I ask them the same question, and they often say they are not following any proven methodology and ask for guidance on what best practices they should follow.
Bottom line, there needs to be greater accountability on board directors and CEOs in their duty of care responsibilities to get data leadership right. AI is all about data, and few companies have effective data lineage, data life cycle management practices locked down, let alone effective and efficient machine learning operations (MLOPs).
I was speaking recently to an EVP of a tier one global bank and was advised that they do not have any centralized AI model libraries to enable researching historical models rapidly to find prior work. Their AI models are not being carefully inventoried, classified, nor do they at present have any formal infrastructure for MLOps, but they are starting to investigate these requirements. Somehow solid AI inventory management practices have slipped through the quality control gaps, perhaps few people understand what really advanced AI and Data Scientist experts really do.
Perhaps we are very naive in making the assumption that AI leaders will learn to classify their models efficiently if we give them more structured tools. Perhaps we simply need to leap ahead and design more Intelligent AI Search Algos to be able to source all connections across the enterprise and be able to know where all the AI models are, be able to classify the users(s) code base, determine its methods automatically and capture any insights derives and do AI inventory MLOP with no human intervention.
We have already learned from years of research in knowledge management and codification methods that humans are consistently messy in organizing their knowledge work.
With knowledge accelerating or the world accumulation of experience which was one to two years at the beginning of the 21st century, fast forward 17 years later, and the doubling of knowledge is now every 12 hours.
So the reality is we cannot wrap our arms around all the knowledge kernals swarming around us, but we do need to build greater depth in these areas to lead and ensure the digital natives have the leadership inspiration to create knowledge clusters to learn from each other.
Understanding how adults learn needs to permeate more the field of AI and stronger leadership and governance programs are needed to advance AI more successfully. At the same time, its critical to balance structure and natural forms of collective intelligence that increase the social ties /nodes of knowledge across the full enterprise – that are now reaching into our home practices as well.