From Agile to AI: Why Culture May Determine the Winners of the AI Revolution

Twenty-five years ago, a small group of software practitioners gathered in the mountains of Utah and produced the Manifesto for Agile Software Development (n.d). Disappointed with the way software projects were developed then, their ideas challenged the dominant project management paradigm of the time: detailed upfront planning, rigid processes, and extensive documentation. Instead, they advocated for adaptability, collaboration, customer feedback, and continuous learning. Little did they really understand the concepts of just-in-time, limiting work in progress, eliminating waste, and progressive elaboration with rolling wave planning that were already foundational to project management where practice created a non-existent theory (Rajagopalan, 2014). 

Although the adaptive Agile ways eventually spread across industries and continents, its adoption was far from uniform. Some organizations embraced it enthusiastically, while others struggled or rejected it altogether. The reasons were often deeper than process or technology; they were cultural. While later Agile practitioners called any Agile Transformation to be a practice rooted in change management to modify the team and organizational culture, the adoption of culture was an after-thought! It shows the myopic foresight and the lack of diversity of the practitioners which is evident from the second manifesto principle "working software over comprehensive documentation" limiting the agile approaches to software development. 

Based on numerous years of research, the Dutch social psychologist Geert Hofstede (1980) provided a comprehensive lens to view the geographical culture across five dimensions first and later adding the sixth dimension. These dimensions include: Power Distance, Individualism, Motivation towards achievement and success (which was originally called Masculinity), Uncertainty Orientation, Long-Term Orientation, and Indulgence (sixth dimension added later). More details of these cultural elements can be found in Hofstede's Culture Factor Group (n.d) or in other scholar-practitioner materials (Hofstede & Bond, 1988; Hofstede, 2011).

In summary, countries with lower power distance often found it easier to embrace self-organizing teams and decentralized decision-making. Cultures with lower uncertainty avoidance were generally more comfortable with experimentation, iterative delivery, and evolving requirements. Highly individualistic societies often adapted quickly to Agile's emphasis on empowered individuals, while collectivist cultures sometimes preferred greater structure and consensus-building. None of these approaches were inherently better or worse, but they influenced how Agile was interpreted, implemented, and sustained. When Agile was promoted because people attended a Scrum Master workshop or gained an online training, little to no attention was given to the cultural considerations of the team members, their hybrid cultures due to long travel, immigration, expatriate arrangements, etc. So, agile didn't fail but we failed agile. 

Author's attempt to compare four countries (May 2026)

Today, organizations face another transformative wave: Artificial Intelligence. Much like Agile, AI is frequently presented as a universal solution capable of improving productivity, accelerating innovation, and transforming business models. The landscape of existing roles are being refined or even redefined where people have began question why do we need developers as AI tools can write them, why do we testers because tools can help with automation, and why do we need project managers, product owners, or scrum masters because AI can replace them! Organizations and leaders have not yet learned the lessons that Agile adoption taught. Extending these observations, AI adoption will not be determined solely by technological capability. As a popular saying goes, "A fool with a tool is still a fool," successful AI adoption needs people to realign AI ways of working with the individuals and team culture, organizational culture, leadership philosophies, risk tolerance, and societal values. The same cultural dimensions that shaped Agile adoption are likely to shape AI adoption as well.

Consider uncertainty avoidance, which is the orientation to risk. Organizations and societies with high uncertainty avoidance often demand predictability, governance, and clear accountability. Such environments may be slower to adopt generative AI systems whose outputs can be probabilistic and occasionally unpredictable. Conversely, cultures more comfortable with ambiguity may experiment aggressively with AI, accepting occasional failures as part of the learning process. Similarly, power distance influences whether AI is viewed as a democratizing force that empowers employees or as a centralized tool controlled by senior leadership. Long-term oriented cultures may prioritize investments in AI capabilities that take years to mature, while short-term oriented cultures may focus on immediate productivity gains and rapid return on investment.

However, AI introduces a dimension that Agile largely avoided: ethics at scale. While Agile transformed how work was organized, AI has the potential to transform how decisions are made. This shift requires leaders to move beyond questions of efficiency and effectiveness toward questions of responsibility and societal impact. Ethical frameworks such as Utilitarianism ask whether AI creates the greatest good for the greatest number. Deontological perspectives ask whether certain actions remain wrong regardless of outcomes. The Indian concept of Lokasamgraha (Singh & Awasthy, 2023) reminds us that decisions should contribute to the welfare and stability of society as a whole. Likewise, the bioethical principles of beneficence, non-maleficence, justice, and autonomy provide valuable guidance for evaluating AI systems that increasingly influence human lives.

One of the most important lessons from Agile is that transformation cannot simply be imposed through frameworks, certifications, or technology investments. Many organizations adopted ceremonies without embracing values, resulting in what practitioners often call "Agile in name only." AI faces a similar risk. Organizations may deploy chatbots, copilots, and predictive algorithms without addressing trust, transparency, governance, workforce readiness, or ethical safeguards. Just as Agile required cultural adaptation rather than procedural compliance, AI adoption requires organizations to align technology with values, leadership behaviors, and stakeholder expectations.

The future may belong not to the organizations that adopt AI the fastest, but to those that adopt it the wisest. Agile taught us that successful transformation occurs when methods align with culture and when people are empowered to learn and adapt. AI extends that challenge from teams and organizations to society itself. As leaders navigate this next era, the question is not simply whether AI can improve performance. The more enduring question is whether our cultural values, ethical principles, and leadership choices can ensure that AI improves humanity as well.

References

The Culture Factor Group (n.d.) Retrieved from https://www.theculturefactor.com/

Hofstede, G. (1980). Culture's Consequences: International Differences in Work-Related Values. Beverly Hills, CA: Sage. 

Hofstede, G. & Bond, M. H. (1988). The Confucius connection: from cultural roots to economic growth. Organizational Dynamics, 16, 4-21.

Hofstede, G. (2011). Dimensionalizing Cultures: The Hofstede Model in Context. Online Readings in Psychology and Culture, 2(1).

Manifesto for Agile Software Development (n.d.). Retrieved from https://agilemanifesto.org/

Rajagopalan, S. (2014). Review of the myths on original software development model. International Journal of Software Engineering & Applications, 5(16), 103-111.

Singh, D. & Awasthy, R. (2023). Lokasamgraha: An indigenous construct for social entrepreneurship. IIMB Management Review, 35(4), 344-358.

Mechanical Empathy: A New Leadership Competency for the AI Era

I have been experimenting with some of the AI tools in developing some new functionality! When I asked it if something can be done, it started doing it! I responded with a prompt, "Don't jump to conclusions dragging me with you! Instead, stop and engage me with powerful questions so that we can collaborate on the solution!" I treated the interaction with the AI tool as a coach would engage with the person coached. Not commanding with prompt engineering as many prompt engineering schools of thought emphasize with the STAR (Situation, Task, Action, Results) method, for instance. During my interactions with AI, I realized the need for "Mechanical Empathy."

The irony of the modern workplace is striking. I have seen people complain, “Gen AI sucks!” In one of the interactions with participants at my talk in Melbourne on "Modern Cost of Quality", people asked, “Why should we worry about workload for robots? They are not living beings.” Yet, I see everyone emphasize empathy, inclusion, psychological safety, and sustainable work practices for human teams. Wait, did we somehow in our rush toward the “AI race,” forget a critical leadership principle: the way we interact with technology shapes the quality of the outcomes we receive. We may not need emotional attachment to machines, but certainly need something deeper and more practical: mechanical empathy. Not empathy for feelings, but empathy for systems, limitations, constraints, context, and capability. As the AI race continues, the more we fail to realize this new leadership competency, the more we will extend that same thoughts on humans. Would that shape a better future?

Consider how we treat children learning to communicate. We do not ask a six-year-old fifty complex questions in thirty seconds and then ridicule them for misspelling a word. We do not hand a new employee the workload of five people and then publicly shame them for underperforming. If we did, most leadership experts would identify the root problem not as incompetence of the child or the employee, but failure in parenting, coaching, management, and system design. Yet this is precisely how many people interact with AI systems today. Users bombard AI with vague prompts, contradictory instructions, overloaded requests, insufficient context, unrealistic expectations, and then critique the results with remarkable confidence. The issue is often not that the machine lacks capability, but that humans lack intentionality in how they engage with it.

Mechanical empathy is not about pretending AI is human. It is about recognizing that every system — human or machine — operates within constraints, patterns, architecture, and context. High-performing leaders already understand this intuitively when working with people. Leadership-level empathy requires presence in the moment, careful observation, appreciation of ideas and situations, curiosity expressed through thoughtful questions, reflection before reaction, risk awareness, ethical consideration, and constructive feedback designed to stimulate improvement rather than humiliation. Ironically, these same leadership behaviors improve AI interactions as well. Clear prompts, structured context, iterative refinement, workload balancing, validation mechanisms, and feedback loops often produce dramatically better outcomes than impulsive criticism. In many ways, effective AI utilization is exposing how poorly many individuals communicate, delegate, and think critically.

There is also a broader organizational implication. As agentic AI systems increasingly become part of the workforce — scheduling work, analyzing data, drafting communications, automating workflows, and even coordinating decisions — organizations may need to rethink the meaning of workforce management itself. Leaders already understand that burned-out human teams make more mistakes under chaotic conditions. Similarly, overloaded AI ecosystems operating with poor governance, weak data quality, fragmented instructions, and unrealistic dependency expectations can amplify errors at scale. Mechanical empathy therefore becomes a form of operational intelligence. It requires leaders to design environments where both humans and machines can perform sustainably and responsibly. The question is no longer simply whether AI is intelligent, but whether humans are interacting intelligently with AI.

Ultimately, the conversation about empathy toward AI is not really about protecting machines. It is about improving humanity’s relationship with technology and reflecting on our own behaviors. The faceless mechanical workforce may not possess emotions, but our interactions with it reveal our patience, discipline, ethics, critical thinking, and leadership maturity. The organizations and individuals who succeed in the AI era will not necessarily be those with the fastest tools, but those who develop the wisdom to engage thoughtfully with both human and mechanical systems. Mechanical empathy may become one of the defining leadership capabilities of the next decade — not because machines demand it, but because effective leadership always begins with understanding the nature, limitations, and potential of the systems we seek to influence.

What are your thoughts? Please share.

AI isn’t replacing leaders - It reveals the leadership gaps

Recently, I took part in a strategy discussion about the need to incorporate AI in designing good practices and preparing the next generation training them on AI tools. I reasoned the focus should not be on execution efficiency but strategic effectiveness. A gentle reset is required understanding the role of AI in future than rushing to adopt AI everywhere! If we look back into the present from a future, is this what we want our next generation to have? 

Artificial Intelligence is often celebrated as the ultimate accelerator of efficiency—automating workflows, optimizing operations, and reducing human error. And to be fair, it delivers on that promise. Organizations today can execute faster, cheaper, and at scale in ways unimaginable a decade ago. But beneath this efficiency boom lies an uncomfortable truth: AI is not closing leadership gaps—it is exposing them. The more we rely on AI for execution, the more visible our shortcomings become in strategic thinking, business acumen, and long-term value creation.

I can certainly say that AI excels at anomaly detection, pattern recognition, and fault prevention, far beyond the mundane automation tasks it is commonly used for. I myself have used AI to create things that would have taken a lot of time! So, yes, it can identify fraud in milliseconds, predict equipment failures before they happen, and surface trends hidden deep within complex datasets. Yet, do these capabilities inherently translate into better strategic decisions? 

Recognizing a pattern is not the same as interpreting its meaning in a volatile market. Detecting anomalies does not equate to understanding their business implications. Drafting even an email does not necessarily connect with the cultural connation of the way the message may be perceived. The gap here is not technological; it is cognitive. Leaders must still ask: Which patterns matter? Which risks are worth taking? Which signals should shape our strategy? AI informs decisions; it does not make them wise. So, do we prepare people for leadership role? Does the use of AI in their responsibilities make someone a leader?

Consider facial recognition technologies. AI systems have reached remarkable levels of accuracy in identifying individuals, yet they continue to struggle with bias. This is not a failure of algorithms alone; it is the lack of risk management thinking leading to the  failure of governance, ethics, and oversight. Bias in AI reflects bias in data, which in turn reflects bias in human systems. Leadership gaps in ethical frameworks, inclusive thinking, and accountability become amplified when scaled through AI. In my mind, AI there is an amplifier of our gaps mainly on leadership level strategic thinking. The question is no longer whether AI can recognize faces, but whether leaders can recognize and correct systemic inequities embedded in their organizations.

Similarly, AI’s prowess in pattern recognition has not guaranteed success in market or product development. Companies have access to unprecedented consumer insights, yet many still fail to create products that resonate or strategies that endure. Why? Because strategy is not just about identifying trends—it is about making choices under uncertainty. It requires judgment, intuition, and the courage to deviate from data when necessary. AI can suggest what is happening, but it cannot define what should happen. That responsibility lies fully with leadership, and it is here that capability gaps—particularly in strategic thinking, customer-centric innovation, process oriented sustainment considerations, alternative impact oriented thinking inherent in risk and people oriented change management become glaringly evident.

Even in highly automated environments like aviation, autopilot systems have not eliminated the need for pilots. Instead, they have elevated the role. Pilots are no longer just operators; they are decision-makers in critical moments when systems fail or unexpected conditions arise. In the healthcare setting, AI enabled systems can identify tumors but have not removed the need for diagnostic image operators, radiologists, physicians, or surgeons. It has only made their role more important. 

The same principle applies to business leadership in the age of AI. As execution becomes increasingly automated, the expectation for leaders shifts toward higher-order capabilities: governance, risk management, ethical judgment, and continuous capability building. AI does not replace leadership and it raises the bar for it. The organizations that will thrive are not those that adopt AI the fastest, but those that close the widening gap between technological capability and strategic leadership maturity.

What are your thoughts? Please comment.

With AI here to stay, what kind of humans will autonomous systems and mechanical robots demand?

I was traveling in India where I had discussions with family and friends around the revolutionary AI landscape. I could sense a feeling of paranoia and confusion. So, I asked myself, "Imagine a future meeting where scheduling is automated, risks are predicted in real time, stakeholder sentiment is analyzed instantly, and portfolio trade-offs are simulated before anyone speaks. The dashboards are perfect." In such a utopian world, what happens when the contextual decisions are problematic and the forecasts are probabilistic. Would the robots turn to the human in the room and say, “Optimization complete. Strategic ambiguity unresolved. Ethical trade-off undefined. Human intervention required.” In my mind, that is not science fiction. That is trajectory.

AI is extraordinarily good at optimization. It can reduce noise in medical images, prevent aircraft drift through autopilot, activate ABS braking systems in milliseconds, and execute trades at speeds no human can match. It detects patterns, flags anomalies, and recommends mitigation paths. But optimization is not direction. Prediction is not purpose. Progress is not value. Algorithms can simulate ten efficient options; they cannot define which future is worth pursuing. They cannot decide what the organization should value when speed conflicts with sustainability, or profit conflicts with reputation.

In such a world, project management does not disappear—it mutates and reemerges. The coordinator of tasks becomes the architect of decisions. The status reporter becomes the framer of ambiguity. The future competency is not mastering more tools; it is mastering judgment. It is rethinking the current process and workflow rather than fall victim to an old tool. It is the ability to think and address risks much before it materializes. It is the ability to define value under uncertainty, to reconcile competing incentives, to make trade-offs that algorithms surface but cannot morally resolve. AI will compress execution layers. What remains, and expands, is decision architecture.

If Agile were written in an AI-native era, it might read differently. Not “responding to change over following a plan,” but conscious human judgment over blind automation. Not velocity metrics over everything else, but strategic intent over algorithmic efficiency. Agile was always about adaptability in complex environments. AI increases complexity. It accelerates data. It amplifies consequence. It does not eliminate the need for leadership—it sharpens it.

The uncomfortable truth is this: AI will not replace project leaders. It will expose those who never moved beyond tools. In a room full of autonomous systems, the only human invited to stay will be the one who can answer: Why are we doing this? Who benefits? What risks are we willing to accept? What future are we choosing? Leadership begins where optimization ends. And in that moment—when the machines pause and wait—the human who can think will matter more than ever.

What are your thoughts?

Has AI killed Agile and Project Management?

Is Agile dead? Do we even need project managers in an AI-driven world? 

These questions surfaced every time technology discussions around AI came up in the last 3-4 months. But before we declare the end of a profession, we must pause and examine a deeper truth: First, AI isn’t new. I was exposed to Expert Systems in 1991-92 when I designed rules based first responder system  using Prolog. If I have exposure to it almost 3 decades back, then, I am sure many others have used it in numerous ways. 

In my experience subsequently, I have found that we have trusted it for years. It already flies planes through auto-pilot systems, prevents skidding through ABS braking, and executes trades in milliseconds through algorithmic platforms. In healthcare, it enhances diagnostic images and flags clinical risks long before the human eye can detect them. Yet in every one of these domains, humans remain accountable. Why? Because context, risk, and ethics cannot be automated. Judgment cannot be outsourced.

The question to ask here is did AI eliminate pilots, drivers, traders, or physicians? No. It elevated them. It removed repetitive execution and exposed the higher-order responsibility of decision-making. The same shift is happening in project management. AI can optimize schedules, analyze risks, summarize meetings, and generate reports. But it cannot align conflicting stakeholders, resolve strategic trade-offs, or lead teams through ambiguity and resistance. It cannot sit in a room where political tension exists, apply context, and choose courage over convenience. It cannot balance short-term delivery pressures with long-term enterprise value. Project management was never about tasks; it was always about decisions.

So, Agile is not dead and neither is project management. What is dying is cargo-cult Agile and checklist-driven project management. Frameworks were never meant to replace thinking but promote it. When we confuse process compliance with leadership, we diminish the profession. The uncomfortable reality is this: AI will not replace project managers, but it will expose those who never learned to think strategically. It will surface who understands value and who only understands velocity. It will reveal who can translate uncertainty into direction and who relies solely on templates.

In Leadership Unleashed, I argue that leadership begins when we move beyond tools and into conscious choice. This moment in history is not a threat to project management; it is a clarifying force. AI optimizes execution. Humans create meaning. AI accelerates data. Leaders shape direction. In a world of increasing complexity, the need is not for fewer project leaders—it is for stronger ones. The future does not belong to those who manage tasks. It belongs to those who can think, decide, and lead when certainty is absent.