Chess has long been considered a metaphor for foresight and strategic decision-making, qualities that are strongly associated with effective leadership. Just as a skilled chess player anticipates moves and manoeuvres pieces for victory, so too does a leader guide her team towards success with the available resources. The 1500-year-old board game continues to offer relevant insights into contemporary management practices and leadership skills.
Chess and AI have a history! It was the first game to witness human versus AI competition. One will recall in 1996, Deep Blue, the AI, played against and lost to the then-reigning world champion, Garry Kasparov, 2.0 to 4.0 points. A year later, in a rematch, Deep Blue beat Garry Kasparov by a solitary win – 3.5-2.5! Within a year, the brute force technique of AI, a statistical method that uses large computing powers to process complex sets of options to eventually decide on one particular outcome/result, understood the nuances of chess even better. AI realised what a robust opening move is. It figured out strategies and realised dominating and defensive positions. It appreciated the context of stalling and sacrificial moves and knew possible outcomes: winning, losing, and drawing a game… without being emotionally connected to any of them. In the passing years, AI has become the unbeatable opposition to human players in chess, and in technology, computing power has scaled up manifold.
AI in the corporate world
In today’s business context, AI is the omnipotent tool that every business leader is contemplating using. The challenge at the leadership level is determining which business activity is best suited for AI deployment that is business-critical and safe. Of course, there’s the lurking fear of failure and job losses for human beings, which also need to be factored in at the strategic level.
When Garry Kasparov was contacted to play against Deep Blue, the players’ fraternity dissuaded Kasparov from accepting the offer. Their fear was that this innovation would kill the game of chess as people would lose interest. However, Garry accepted, played, and history was created in 1997 when AI defeated the champion. Since then, the interest in chess has only increased. Children and intermediary players are playing chess online! As for the experts, they have started to value the virtues of AI as a trainer, an assistant, and even an advisor. AI in chess has created a revolution!
A similar experience is envisaged in the business world with AI’s integration into business activities. The Large Language Model (LLM) is likened to AI assisting a chess player to strategise, anticipate, and organise resources in an optimum manner. LLM offers broad indicators and predictive analytics for leaders to make informed decisions. Although their accuracy may be questionable, their ability to provide a quick start is just unimaginable. Like in chess, LLM will learn and become more accurate and proficient with time.
The world of business changed on 30 November 2022, when people experienced Chat GPT! The tech world is busy building models that are more reliable and human-friendly, meaning human beings understand how AI has evaluated various options, thus making LLM a very compelling companion in day-to-day activities. Hence, AI is getting into every gadget, device, system, and form of interaction – be it the intelligent web browser, interactive app on the mobile phone, or chatbot interacting with a human being on a business call!
Imagining the future
Let’s get a bit creative and gaze through the crystal ball of AI – the omnipresent, intelligent agent that can be the know-it-all ‘boss’ and the diligent, reliable, efficient ‘reporter’ at the same time!
As computing power increases and its access cost reduces, AI will become the central force that drives all activities, including imagination! So, imagine the chessboard being AI-enabled. The board now has its intelligence with the ability to understand the context of the game to prompt the next set of moves. The difference between the board-level AI and the AI used by the player as her assistant is that the assistant knows the player’s psyche of defending or attacking, strengths and weaknesses of the player and her opponent, and factors these while offering suggestions. The two AIs may or may not be aligned in their suggestions since both may be accessing different references.
Let’s activate the third dimension in chess – the pieces are also intelligent! They know their roles and those of the others. They too can think, strategise, and suggest. For instance, in a choice to move between the rook and the knight, the rook suggests the knight moves. The knight feels the Queen should move! This is the egalitarian version of chess! Does it feel real and practical?
In the context of AI, there’s the Large Language Model, which processes data from a vast set of sources with a large number of constraints and rules. It takes a holistic approach, broad-bases its responses, which can be fine-tuned through an iterative process, and thrives on the complexity of the challenge. It also uses extensive computing power. This is where Generative AI (GenAI) tools like ChatGPT, Gemini, and Claude are popular. They are responsive, fast, reasonably accurate, understand the context and the chain of interactions, and also learn from and during the interactions. They are topic-agnostic – we can interact with GenAI on a wide variety of disparate subjects.
The AI world is also moving towards Narrow Language Models and Small Language Models (SLM), which are likened to the chessboard’s AI. SLM is in contrast to LLM – it has fewer data points and constraints and uses less computing power. Hence, SLM is a good tool for subject-specific interactions, which can be reduced to the level of a company’s knowledge base – bringing intelligence to functions like planning, forecasting, budgeting, helpdesk management, etc. Narrow Language Models are like domain experts – they focus on industry verticals and specialise in specific areas.
Narrow and Small Language Models have the potential to transform existing business processes into intelligent business processes. Imagine order-to-cash, procure-to-pay, and hire-to-retire kinds of extensive business processes being managed by systems without human intervention! Or, a legal AI system that can churn out agreements in a jiffy based on the specific contexts of an engagement.
AI and leadership
As one hears about and sees AI everywhere, there appears a strong possibility of the three dimensions of chess becoming a reality. There will be an LLM to provide general guidance. The SLM and Narrow Language will work on specific areas of complexity and repeatability with high reliability, which significantly speeds up operations. AI will also suggest options for decision-making for human beings to act on them. Finally, the organisation’s intelligent workforce will also participate in the decision-making process. One expects this to be very different from current practices when AI becomes active in an organisation’s operations. The command and control hierarchy, which equates accountability to the authority of decision-making, will dramatically change.
The accelerated speed of operations enabled by AI will also demand changes in organisation structure, roles, responsibilities, and functions that human beings will perform. The speed of change and the presence of AI will also induce a higher order of complexity, which will require multi-dimensional perspectives and multi-disciplinary expertise for effective decision-making and building trust in leadership! This is akin to the human player not being able to clearly articulate a decision of instinct or gut – a very human reaction to an external stimulus – while playing the game. The player may be able to defend or justify that rationale only after the game is over. Leadership will face such inexplicable situations more often.
In this phase of evolution, where AI assists a human chess player, the environment tends to praise the AI for the win and blame the human for the loss. When the game is in progress, the total responsibility of decision-making is on the human being. Going forward, leadership will face similar situations. These are new domains of testing leadership skills, which need to strike a fine balance between AI’s intelligence and human instinct; deftly navigate through external and internal sources of information when they differ in their viewpoints; the ability to collaborate with multiple entities of people and processes; allow one’s intuition and instinct to express to take bold decisions when confronted with unfamiliar situations that machines may not have expertise in; be tactful, foreseeing, responsive, and wise while explaining the rationale of a decision, especially when it deviates from any of the recommendations made by all the AI systems in use. Leadership is also about ensuring and explaining how activities have been conducted ethically, keeping the best interest of society. It is also about sensing and managing biases, an inherent human feature, which can now manifest in AI systems. Leadership is about guiding to success the combined wisdom of human beings and AI!