AI is recoding leadership in the Information Age

Speaking as a management educator, one very real problem we face is that the management playbook was written over a hundred years ago. The kind of leaders that were envisaged at that time would hardly fit the bill today. Management theory was created for the Industrial Age, but we live in an Information Age, dominated by data and AI. 

Computational technology, including AI (in all of its forms) has disrupted traditional hierarchies , which are fundamentally built on informational asymmetries. The managerial principles of planning, organising, controlling and leading in an environment which is volatile, uncertain, complex and ambiguous (VUCA) start to break down.  

AI is a powerful socio-technological tool that influences how our society is led. I will outline the formation of traditional notions of leadership and how this is changing. In the Information Age, a new type of leadership is emerging, which I will outline and add the challenges this creates. Ultimately, I will show how the AI aware leader is less a choice and more of a necessity.  

Old School Leadership 

Great thinkers, such as Plato mused about the essential traits of effective leaders millennia ago but the Industrial Age ushered in the principles of workforce management with which we are so familiar today. Ironically, these principles have come to be considered as traditional or old school leadership. They are anything but.  

For the vast majority of human existence we lived in small mobile bands where leadership, as we know it, did not exist. From what we can gather (from modern hunter-gatherers and small-scale societies) leadership was distributed and barely noticeable. Groups were highly collaborative and influence (rather than leadership) changed hands, depending on task. 

What made this change? The increasing stratification and complexity of society required an increasing level of coordination, while wealth inequalities gave individuals power over others. By the age of steam, workforces were large and highly specialised. A new craft, that of management, emerged to coordinate and strategise for increasingly large organisations. 

Despite technological breakthroughs, information technology took longer to develop. Managers had to coordinate sophisticated projects while information was scarce. This led to the, now, familiar hierarchical or cyclical command structure. Information cycles through the structure to the decision-makers and back to the front-line. This was an age of efficiency. 

The Rise of Data and AI 

Towards the end of the 19th Century, information technology and data began to impact our societies, these were to rupture the cyclical decision making processes that had guided the Industrial Age. The first wave of big data occurred in 1890 when the Automatic Tabulating Machine was designed to collect data during the US Census. 

It was no coincidence that the machine was designed by Hermann Hollerith, who founded International Business Machines (IBM) in 1924. Computers were designed to tame the increasing volume of data the world was experiencing, but ultimately they generated more data than humans could. This created a cycle of innovation that would result in the birth of AI. 

AI bursts onto the scene far earlier than many realise. The term Artificial Intelligence was coined by John McCarthy in 1956 at the famous Dartmouth Workshop, the first AI research gathering, attended by mathematicians such as McCarthy, Shannon and Simon. A few years before this, Alan Turing had pondered whether ‘computers can think?’ 

Whether or not they can, the Golden age of AI followed between 1956 and 1974, a period of unprecedented growth in AI capacity, the likes of which has never repeated. Through cycles of AI Summer and Winters, AI has grown haltingly for 70 years into the generative language driven models of today. It is the ability of AI to connect data that fundamentally changes how we lead. 

Leadership and Information 

Leadership is an information game. Short of physical coercion, leaders rely on informational asymmetries to gain influence. Leaders show the way. In the information scarce world of the Industrial Age, knowledgeable leaders called the shots. 

But information technology, computation, data and then AI changed all this. Now information is not scarce. Perhaps there is too much of the damn stuff? The cyclical process of decision making no longer make sense when employees on the ground have access to the same information as the bosses, they can respond to change immediately. It is the age of agility. 

Agile organisations often use decentralised structures. An early example is the Battle of Trafalgar (1805). Admiral Lord Nelson’s, tactical unorthodoxy was his inclination to share his battle plans with his officers. This highly irregular approach allowed his Sea Captains to be autonomous and agile to devastating effect, while their enemies remained inflexible.  

What has this to do with AI? AI allows access to data which can turn us all into agile operators, if we know how to use it. In our modern networked world, cyclical decision making structures are not dead, but they are scarcer. In recent research I conducted in the Royal Navy, I showed how modern teams often use both formal and informal networks to make decisions. 

Good, old fashioned, Command and Control, still has its place. If you find yourself in a high consequence and dynamic situation, you will likely be assured when Emergency Operating Procedures (EOPS) kick in and everyone knows exactly what to do. But increasingly, situations are complex and ambiguous. Does AI help us here? 

Conditions for Information Age Leadership 

Clearly, AI is not a silver bullet. It cannot replace leadership and it arguably adds to the Volatile, Uncertain, Complex and Ambiguous (VUCA) conditions of the 21st Century. My argument is not for AI-assisted leadership. Like any technology, instrumentalism and hype only cloud the picture. Leaders are leaders regardless of the tools they use. 

My argument is not that AI affects leaders directly, but it shapes the conditions of leadership. Decentralised teams rely primarily on trust to operate, and this, therefore, has to be the prime concern of the Information Age Leader. Trust in people, sure, but also trust in the data and the AI technology. Now this is not easy. 

Trust in AI comes not from the technology itself, as we know genAI (specifically) is not designed to answer questions accurately, just plausibly. This is not enough for trust, so teams have to trust they can integrate and use AI effectively and trust in their own skills and knowledge. Trust harnesses AI, but it is something you have to work at. 

To trust your ability to use AI effectively, ethically and responsibly, you have to learn how to use the technology to help you create knowledge. Since only humans can create knowledge, leaders allow AI to organise data and generate information, while supporting humans to assign meaning and test for truth. 

Challenges for Information Age Leaders 

The challenges of using AI in decision making are well documented. Since knowledge must be true and valid for the problem, there is considerable work to be done in turning AI retrieved data into information and knowledge. This is the primary ethical service the leader can deliver, since misinformation is the greatest threat to the Information Age Leader. 

Ethical dilemmas emerge from unequal access to and ability to use AI tools, the inherent biases, and climate damage. AI is an energy intensive industry, this fact accounts for the halting progress of AI since the 1950’s. Complexity barriers require ever larger amounts of energy to leap. An ethical question is how to balance energy expenditure against benefits.  

The AI Aware Leader 

Frank Herbert, creator of Dune, imagined that AI once developed would quickly be purged from human society. He was not the only famous science fiction writer to make this assumption. We would appear to be standing at the beginning of such an arc, the point where (as Roy Amara warns us) our expectations of AI are unduly high. 

Amara’s law also tells us that we will underestimate AI in the long term. In its ability to connect data, AI effectively short cuts the hierarchical model which underpins ‘old school leadership’. One of the unanticipated long-term impacts may be to recode leadership to a point where flat structures become the norm. 

As in any paradigm, too much of one thing is never good. In the same way that we may tire of hierarchy, agile but noisy networks can also be exhausting. They may not be the right fit for all situations. The Information Age Leader straddles this rupture in leadership traditions, and she has no easy tangle of data, information, misinformation and knowledge to untie.