I am tired of hearing AI’s enormous price tag treated as proof that the technology has failed. It has not. Every time someone points at a billion-dollar data center and says, “See? No return,” I want to remind them what we are actually buying. We are not buying a slightly faster spreadsheet. We are paying the upfront cost of a transformation that will change how all work gets done. And yes, that cost hurts. It shows up in quarterly earnings, in capital budgets, in layoffs that make headlines. But the real danger isn’t spending too much today. The real danger is refusing to change while everyone else learns how to work differently. I understand why people flinch. AI is expensive, confusing, sometimes wrong, and often scary. I have sat in boardrooms where the CFO winced at the price of a single model rollout. I have heard executives say, “This better pay for itself by next year.” I get it. But the cost of standing still is much higher, and it compounds silently. The companies that hesitate today will not avoid disruption; they will simply experience it later, when they are forced to catch up from a position of weakness. We need to stop treating AI as a vendor purchase and start treating it as a generational investment. There is no version of the future where we look back and say, “I’m glad we saved the money by ignoring intelligence that can think and create and analyze at machine speed.” That is not how technology works. That has never been how technology works. I have seen this story before. I remember being told that the internet was a fad, that mobile phones were a toy, that cloud computing would never be secure enough for serious companies. Every one of those predictions was confident. Every one of them was wrong.
I lived through the dot-com era. I remember when businesses spent millions building websites, e-commerce operations, and digital marketing systems before the returns were obvious. Some of those investments were waste. Some failed spectacularly. Others took years to mature and required endless patience, false starts, and red ink. But the companies that stayed with the transformation—the ones that kept learning, kept iterating, kept investing in digital infrastructure even when the skeptics were loud—were the ones that changed their industries. Amazon is the example everyone uses, because it survived the bubble and then remade retail. It did not happen because Amazon was smarter or luckier. It happened because Amazon treated digital transformation as a long-term bet rather than a quarterly cost. I remember people saying in 1998, “Why do we need a website? Nobody is buying anything online.” They were right, in that moment. But the companies that built websites anyway learned something every day. They learned what worked, what customers wanted, and how to correct mistakes. By the time e-commerce became normal, they were already proficient. The companies that waited were behind. AI is entering that same uncomfortable middle ground. Forbes reported that more than 142,000 technology jobs were eliminated in the first five months of 2026 while major technology companies committed roughly $700 billion to AI infrastructure. Those numbers raise a legitimate and painful question: Are companies cutting people because AI can already do their jobs, or because leaders believe AI eventually will? Those are two very different decisions, and they lead to very different outcomes. If you fire people because you have a hunch that AI will replace them someday, you are destroying your talent pipeline before your technology is ready to carry the load. If you fire people because you have carefully mapped the work, tested the systems, measured the performance, and proven that the job can be done safely and consistently by AI, that is a harder but more defensible decision. We need to be honest about which one we are doing. Workers are worried about disappearing jobs. Young people are worried that the first rung of the career ladder is disappearing. Those concerns are real, and they should not be dismissed.
That is why I reject the idea that the answer is to stop investing in AI. The answer is to invest intelligently and bring people through the transition. Here is where companies are getting into trouble. AI is excellent at routine work, research, drafting, summarizing, administrative processes, and other defined tasks. It can scan thousands of documents, identify patterns, and produce a credible first draft faster than any human. It can automate the busywork that once consumed entire departments. But it becomes less reliable when it has to make decisions in a Volatile, Uncertain, Complex, and Ambiguous world—a VUCA world. A model can learn the normal cases. It can master the patterns that appear in training data. But business leaders still have to handle the exceptions, the incomplete information, the shifting customers, the ethical dilemmas, and the problems nobody saw coming. AI cannot do that. It cannot look at a long-time client and sense that something has changed. It cannot read the room in a contentious negotiation. It cannot take responsibility for a decision that will affect hundreds of families. I think about a supply chain manager who can see a disruption coming because a supplier is behaving strangely. AI might flag a late order, but it cannot call the supplier’s trusted contact and hear the nervousness in their voice. It cannot decide whether to expedite a shipment or redesign the product. Those are human decisions. That is why firing people before AI can actually perform their entire jobs is backward. It treats AI as a replacement when it should be a support system. Use AI to remove repetitive work. Give people more time for judgment, innovation, customer relationships, and real problem-solving. Keep humans in the foreground. Put AI in the background, where it can build capacity without erasing humanity. This is not just a philosophical position. It is practical. The value of AI is not in eliminating people. The value is in freeing people to do the things that create genuine competitive advantage. The companies that understand this will not be the ones with the most impressive AI systems. They will be the ones whose people are strongest, most curious, and most capable of navigating the unexpected.
The numbers point in the same direction. According to KPMG’s June 2026 Global AI Pulse, only 7 percent of leaders reported established AI return on investment. That sounds damning, but look closer. Organizations with CEO accountability for AI-informed decisions reported established ROI at 14 percent, compared with 4 percent among organizations without that accountability. And leaders with strong cost visibility were five times more likely to report ROI. It would be easy to read that 7 percent and conclude that AI is all hype. But the same report shows that accountability and visibility change everything. The lesson is not “AI doesn’t work.” The lesson is “unmanaged AI doesn’t work.” What does that tell me? It tells me AI is not a magic switch. It is a management challenge. The companies that treat AI as a toy, a bolt-on, or a way to look modern will not see returns. The companies that treat AI with discipline—with ownership, measurement, and clear accountability—will. That is not evidence that AI has failed. It is evidence that we are failing to lead it. I have seen this pattern in every major technology shift. The first wave is always chaotic. People buy too much, promise too much, and then get disappointed when the world does not instantly change. Then the second wave happens, quietly, among the companies that kept working. They install the right governance. They train their people. They set realistic goals. They adjust. They measure. And suddenly, the returns show up. We are already starting to see that pattern with AI. The difference between the 7 percent and the 14 percent is not luck. It is leadership. It is someone in the organization saying, “I own this. I will measure this. I will make sure this investment produces value.” Every company should hear that message. If you are not seeing AI ROI, the first question should not be “Does AI work?” It should be “Who is accountable for making it work?” And after that, “What is the plan, what are the measurements, and how will we course-correct?” That is how every worthwhile investment has ever been managed.
That discipline must reach education, and it must reach the individual level. We cannot keep sending graduates into a workplace designed around entry-level tasks that machines increasingly perform. This is one of the most heartbreaking and avoidable failures of our time. Young people do everything right. They go to college. They get degrees. They follow the traditional path. And then they discover that the entry-level job that used to be the first step—the research assistant, the junior analyst, the administrative coordinator—has been automated or restructured. We owe them more than advice like “learn to code.” We need to redesign education around what human beings do better than machines. Students need specialized skills, yes. But they also need adaptability, critical thinking, judgment, creativity, and the ability to work alongside AI from day one. AI will be their coworker. They need to know how to assign work, check work, and correct work. They need to know what AI is good at and what it is not. They need to practice making decisions with incomplete information. They need to learn how to communicate, persuade, and build trust in a world where anyone can generate a perfect email. Employers also need to rethink entry-level roles. Those roles should not be treated as disposable labor or a cheap way to get work done. They should be training grounds for uniquely human capabilities—the judgment, the relationships, the intuition, the responsibility that eventually carries an organization forward. Individuals have a choice, too. I understand the fear. It is real. It is in every email I get from someone who asks, “Will I have a job in five years?” But fear can freeze you, or it can force you to adapt. The people who come out ahead in this transition will not be the ones who hide from AI. They will be the ones who learn the tools, understand their industry, and build the skills AI struggles to replicate: trust, persuasion, creativity, leadership, relationship-building, and judgment under uncertainty. That is not an easy list. But it is not impossible, either. It is the oldest kind of value, the kind that has always been at the center of human work.
Business leaders must make the same choice. Do not treat AI as a replacement program. Do not treat it as a software purchase that you can install and forget. Treat it as a new member of the workforce. Define its responsibilities. Set boundaries. Train your people to work with it, not around it. Measure what it saves and what it produces. Keep human beings accountable for consequential decisions. That means no AI-generated decision should be an unattributed decision. If something goes wrong, a human must be able to explain what happened and why. If we do this well, the payoff is enormous: lower routine costs, faster learning, stronger customer relationships, more productive workers, and organizations capable of doing far more with the same resources. If we do it badly, we get the opposite: hollowed-out talent pipelines, frightened workers, expensive systems nobody trusts, and companies that discover too late that automation cannot replace institutional knowledge, loyalty, or the ability to navigate the unexpected. We are already living in a VUCA world. Our organizations and education systems must catch up. That means investing in AI. It also means investing in people. It means measuring results, telling the truth about what works and what does not, and redesigning work around what humans and machines each do best. The companies that make that transition will build the future. The ones that refuse will eventually have to buy their way back into it—and they will pay more for it then than they would have paid today. This is not a time to panic. It is not a time to retreat. It is a time to be honest about what transformation costs and why the cost is worth paying. We have been through this before. Tired as it is, costly as it is, scary as it is, this is how progress has always worked. The choice is not whether to change. The choice is whether we change wisely and compassionately, or whether we stumble into the future with our eyes closed.
