When I think about artificial intelligence, I often find myself thinking about the horse. Most of us give very little thought to what it took to keep a city moving when horses were the principal means of transportation. They needed food and care, their manure had to be cleared from the streets, and animals that died from exhaustion had to be removed without the powered equipment we take for granted. An enormous amount of work went into sustaining that system. It provided livelihoods, and it also imposed hardships that are easy to overlook from a distance.

The arrival of the railroad brought its own disruption. A steam engine belching smoke and a railway cutting across someone’s property hardly looked like an unqualified improvement, particularly to people whose businesses depended on horses. Yet think about what it meant to make a long journey by train instead of a horse-drawn carriage or an ox-drawn wagon. That change opened possibilities for travel and commerce that could not be understood simply by counting the occupations it threatened.

As a historian, I find that perspective useful when thinking about AI. Earlier generations lived through changes that unsettled their understanding of work and the future, and generations after us will do the same. As the pace of technological change accelerates, leaders have less time to understand new possibilities and help their organizations respond. I approach that challenge with considerable optimism. History gives us reasons for confidence, along with examples of how a leader’s willingness to learn and sense of responsibility can shape the outcome.

Look beyond the first disruption

The railroad story becomes even more interesting when you follow what grew around it. A railway needed people who could build and maintain locomotives, and its expansion helped stimulate industries such as steel. Running trains also required communication over long distances. Telegraph operators became part of the work of coordinating their movements, and railway schedules made the differences between local times increasingly troublesome. A town could manage perfectly well by its own clock until people and trains needed to move reliably between towns. Railways helped drive the adoption of standardized time, linking transportation to developments in communication and timekeeping.

I think that is one of the most useful lessons for leaders considering AI. A new capability creates needs that are difficult to see before people begin using it. Someone looking only at the work displaced by the locomotive would have missed the occupations and businesses that developed around the railway. I believe AI can open the way to new jobs, better pay, and more satisfying work as people discover what they can accomplish with it. For a business leader, that means looking closely at the problems clients face and helping them find solutions that new capabilities make possible. Leaders need to give their people room to explore those possibilities.

I also think we should be honest about the role of personal gain. The people who built new industries often wanted to succeed financially, and many also wanted to make life better for others. Those ambitions can reinforce one another. With AI, I see an opportunity to build a stronger business by helping people do more valuable work and finding better ways to meet their needs. Leaders should keep that connection in view as they decide where to invest and what success should look like.

Learn from others and act on what you learn

Leaders can benefit from the investments and hard lessons of others if they are willing to change how their own organizations work. Think about Britain and Germany in the nineteenth century. Britain’s first Industrial Revolution, roughly 1760 to 1840, had given its manufacturers decades of experience with textile mills and steam-powered production. Germany’s rapid industrial expansion gathered pace in the 1840s and 1850s and continued through the decades before World War I in 1914. A new German enterprise could draw on technology and experience developed in Britain and elsewhere, then build around more recent equipment and methods. An established manufacturer faced the cost of replacing machinery and changing routines already in place.

The established manufacturer could look at its machinery and say, “It’s good enough. It’s still making money for us.” Replacing it meant spending money and changing the way people worked, so postponing the decision could seem perfectly reasonable. Yet a competitor using newer methods might steadily become more productive. For me, the lesson is that an organization can fall behind while its existing approach still appears to work. With AI, leaders need to learn from what others are achieving and be willing to reconsider familiar processes before the gap becomes difficult to close.

That helps explain why a nimble startup can sometimes move faster than a much larger competitor. It can learn from the investments and mistakes of established businesses, then apply those lessons without having as much tied up in older systems and ways of working. A handful of people can decide to try something and change direction quickly. Larger organizations have resources and experience that give them real advantages. The leadership challenge is to use those strengths while keeping the organization open to new ideas and able to act on them.

That also gives a smaller company something to think about as it grows. Size brings people and resources that allow an organization to do more, but it can also bring bureaucracy, friction between teams, and a loss of alignment. I would keep asking whether we are actually getting better at what we do. If we are adding people and systems, those additions should improve what we can accomplish. Getting bigger for its own sake can cost us the very qualities that allowed us to make progress.

I saw the value of learning from outside an institution during my Army career. When I was a battalion commander in Iraq, we used commercial cell phones and publicly available mapping tools because they were often better than the alternatives available to us. We still had real requirements and safeguards to consider. For me, that experience illustrates the openness leaders need when adopting AI: recognize useful capabilities beyond established systems and judge how they can be used responsibly.

Keep adapting after you succeed

Henry Ford shows how a leader can help bring about a profound change and then struggle to respond to the next one. Ford and his engineers drew on methods used in other industries as they developed the moving automobile assembly line. Moving a car through a sequence of tasks allowed workers to concentrate on particular steps, helping Ford produce cars more quickly and at lower cost. The result changed what people could afford and helped create a much larger market for automobiles.

Then success created a different challenge. Ford became deeply attached to the Model T and continued concentrating on its cost and reliability as buyers began to want greater comfort and style. General Motors responded to those changing expectations with refreshed models and more choice. Ford’s manufacturing accomplishment remained extraordinary, but the market was moving beyond the assumptions that had made it successful. That is a familiar danger for leaders. When an approach has served us exceptionally well, it becomes easier to defend it and harder to recognize when the people we serve are asking for something different.

The Wright brothers offer a related lesson. Their work on controlling an aircraft helped make practical flight possible, and they had every reason to take pride in what they had achieved. Yet patent disputes drew an enormous amount of their attention into defending that achievement, while other pioneers, including Glenn Curtiss and aviators in France, continued to advance. What strikes me is how inwardly focused they became as the possibilities of flight were opening up. Their experience reminds me to keep collaborating and learning from others after a success. For leaders adopting AI, an early breakthrough should encourage continued experimentation and a willingness to draw on ideas beyond our own organization.

Keep human judgment at the center

I am of a certain age, so I remember the concern when calculators became widely available. People worried that they would ruin young minds and undermine the teaching of mathematics. Used properly, they allowed students to go much further because they saved time on calculations. A student could spend more effort understanding a difficult problem and exploring how to solve it. The educational responsibility was still there: students needed to understand what they were doing well enough to recognize an answer that made no sense.

I approach AI in much the same way. It is our tool, and we remain responsible for how we use it. During my years educating future officers at West Point, developing judgment was central to the work. An AI-generated answer still needs someone who knows the subject to examine it. That person may find an outright mistake, or something subtler: a lack of nuance, a misplaced emphasis, or an answer that fails to address the real problem. The ability to notice those shortcomings comes from understanding the work and its purpose.

AI also gives leaders a reason to make better use of experience. I think of scholars who produce some of their most important work in their seventies and eighties, drawing on decades of study. Organizations need new people and fresh thinking, and they should also give experienced colleagues the opportunity to continue doing important work. Their role can include evaluating AI’s output and helping others learn, while also putting their own knowledge to work on questions that matter. We should be looking for ways to keep that experience contributing to the organization.

Build the trust that allows people to follow

For me, leadership begins with character. We want competence and courage in the people we follow, and we want to know that they understand the situation and care about us. Trust develops as we come to know them and see how they exercise responsibility. AI cannot create that relationship for us. We have to earn that confidence ourselves, with our colleagues and with our clients. As organizations change, people will want to know what the change means for them and whether they can trust the person leading it.

I also draw a distinction between a great innovator and a leader I would want others to emulate. Someone may see a possibility that everyone else has missed and create something extraordinary. Building a team and carrying an organization forward call for qualities that deserve their own attention. When we study successful people, we should look at how they lead others as carefully as we look at what they invented. That matters for AI because the technology will be only part of what an organization needs to make progress.

I see the need for human connection in my work developing leaders. Bringing people together gives them a chance to understand one another and build relationships that help them work through difficult decisions. A leader can have extraordinary information at hand and still need to sit with a team, hear its concerns, and explain a judgment. The same is true of our clients: they need to know that we understand their situation and care about getting it right. As AI helps us work more quickly, I believe those human relationships will become even more necessary.

Take responsibility for the transition

My optimism does not lead me to discount fear and anxiety. The Luddites of the early nineteenth century were textile workers whose livelihoods were threatened by changes in production. Some responded by destroying machinery, hoping to protect the value of their skills and their ability to earn a living. Their fears had a real basis, and industrialization continued despite their resistance. I believe AI’s benefits will likewise keep drawing people toward it. Leaders need to engage with that change and help their people find a way forward. Telling someone about a better future offers little comfort if they cannot see how they will have a place in it.

The history of industrialization also includes child labor, dangerous working conditions, and serious pollution. Those harms demanded action, and the experience of earlier transitions should help us recognize and address problems sooner. With AI, leaders need to understand what they are introducing and take sensible measures to prevent abuses and correct mistakes. They also need to help people learn how their work can develop. Confidence in the opportunity gives us a reason to take those responsibilities seriously and to act on problems as they become apparent.

I believe AI can free up the best in humanity to accomplish things we cannot yet imagine. When I think about what earlier advances made possible, I think about better medical care, wider access to education, and more time to do things beyond the work of getting through the day. We are participating in that continuing history, with less time to absorb each new development. I have great confidence in the human mind and in human goodwill. Our task as leaders is to put that confidence into practice, helping people adapt and earning their trust as we move forward together.