By Melvin Bosso

My name is Annie Archer, and I drive the bus from Abidjan to Cotonou. It is a long journey along the coast, nearly eight hours when the roads, borders, traffic, and weather cooperate. Long enough for passengers to settle in, open their laptops, take calls, and eventually begin discussing the things changing their lives.

Lately, one subject returns more than any other: artificial intelligence.

It is 2026. We are at least six years into AI’s modern rise and roughly two years into an acceleration that feels unmistakably exponential. What began as an emerging technology has become a defining economic force. AI investment has helped fuel recent growth in the United States, while governments, investors, and companies elsewhere are moving quickly to avoid being left behind. Across industries and economies, the message is clear: artificial intelligence is no longer an interesting experiment. It is becoming a strategic priority.

Yet it is easy to misunderstand what this change represents. Many still see AI as another technology rollout: another platform, another piece of software, another initiative to implement. But AI is doing something more fundamental. It is changing our relationship with information, how quickly we access it, collect it, review it, structure it, analyse it, and increasingly turn it into action.

In that sense, AI may become to the office worker what robotics became to the line operator. The technologies are different, and so are their consequences. Still, the comparison helps us understand the scale of the shift. Industrial robots transformed repetitive physical work. AI is beginning to transform repetitive cognitive work: preparing presentations, reconciling information, drafting communications, summarizing reports, identifying patterns, and supporting decisions.

That transformation brings opportunity. It can help people work faster, improve execution, reduce errors, and focus on higher-value tasks. But it also raises questions about jobs, skills, investment, energy, regulation, social trust, and the long-term role of human judgment. Those questions are no longer theoretical. They are shaping boardroom conversations, operating models, government policy, education systems, and the daily experience of employees.

I have listened to enough people on this bus to know that there is no single view of AI. Some passengers see liberation: more time, creativity, and output. Others see uncertainty: new tools arriving faster than organizations can understand or absorb them. Some see a strategic race; others see an economic bubble. Some see an extraordinary technical breakthrough; others see a social transformation whose consequences may take generations to understand.

The most pessimistic observer may imagine AI’s path as a smooth climb leading to a cliff, a singular edge beyond which the technology, its economics, or its social acceptance simply fall away. It is a powerful image, but perhaps not the most useful one. History suggests that major revolutions rarely end at one sudden edge. Steam power, electricity, the automobile, and the internet did not move in straight lines. They advanced through enthusiasm, resistance, correction, adaptation, and reinvention.

AI is likely to follow the same pattern. It is not necessarily moving toward a cliff. It is moving toward a series of tipping points: moments when tensions across technology, energy, economics, regulation, and society converge and force a change in direction. A tipping point does not necessarily mean collapse. It can mean recalibration. It can create new constraints, business models, rules, and opportunities.

Understanding those tipping points matters to everyone whose career, capital, organization, or future depends on AI: the employee learning to work alongside it, the manager trying to introduce it responsibly, the strategist deciding where to invest, the CEO accountable for returns, the expert concerned about safety, and the educator wondering what tomorrow’s leaders will need to know.

On this journey from Abidjan to Cotonou, six passengers give voice to those perspectives. Each sees AI from a different seat. Each carries a different concern, expectation, or ambition. Together, their stories offer a rounded view of artificial intelligence, not as one technology with one future, but as a force reshaping how we work, invest, lead, learn, and live.

I am Annie Archer. I drive the bus. These are the conversations I have been hearing.

Jim, the Practitioner: “I Can Do More With So Much Less”

Jim has the window seat and the relaxed posture of someone whose work has genuinely improved. An analyst by trade, he has watched AI take over the parts of his job that once consumed time without adding much value: preparing first drafts of presentations, pulling data from multiple sources, reconciling spreadsheets that never quite aligned, and formatting documents for review.

“Thanks to AI, I can execute what I need to execute at a quicker pace,” Jim says as palm trees blur past the window. “I feel like I can do more with less.”

For Jim, AI has not replaced his role; it has changed its nature. He no longer spends his best hours completing small, repetitive tasks. Instead, he uses more time to shape ideas, assess information, solve problems, and support decisions. Presentation analysis, first-pass synthesis, document formatting, and routine research no longer stand between him and meaningful work. They have become the starting point rather than the destination.

The change is not only about speed. It is also about attention. Jim can move from collecting information to interpreting it. He spends less time producing material and more time asking whether it is useful. He can develop stronger recommendations, test ideas more quickly, and focus on questions that require judgment, context, and experience.

His adoption of AI happened gradually, one workflow at a time. There was no single day when his work changed completely. It began with a first draft, then a presentation, then a data analysis, and finally the realization that he could complete in hours what might once have taken days. His personal tipping point was quiet, but decisive.

Near the end of the journey, Diamond turns toward him.

“Jim, would you go back to how things were before AI?”

Jim does not hesitate.

“No. I would not.”

Diamond pauses before asking a second question.

“What is the single most important benefit of AI, according to you?”

Jim looks out of the window for a moment before answering.

“Time. It gives me time, time to think, create, solve more meaningful problems, and do more with less.”

Alexandra, the Middle Manager: “My Brain, My Heart Are Open”

Alexandra sits two rows behind Jim, her laptop open as she scrolls through emails, meeting notes, and a growing list of AI-enabled tools someone has asked her to explore. She is a middle manager, and her relationship with AI is perhaps the most complicated on the bus. She is neither resistant nor entirely at ease. She sees the opportunity, but she also carries the uncertainty of being responsible for a team while the nature of work changes around her.

“I have AI coming at me from all sides,” Alexandra says. “I can see how it is transforming my team’s job. I can see how it is going to influence my own job, especially on the execution side. But it is also forcing me to think differently.”

Her experience is far from isolated. Forty-seven percent of middle managers and individual contributors report experiencing at least one negative effect from AI adoption, compared with 31% of executives and senior leaders. Those closest to implementation are often the ones feeling the pressure most directly.

“I do not always know what is available for me to introduce to my team,” she continues. “I do not always know how my organization is thinking about AI, what the priorities are, or where the limits should be. But I want to learn what I need to learn. I am ready. My brain is open. I am happy to integrate it. I have started using AI-enabled software, but I know there is more to come. So, give it to me. Let me work through it.”

Alexandra lives in the space between strategy and execution. Senior leaders speak about transformation, productivity, efficiency, and competitive advantage. Her team wants practical guidance. They want to know which tools they should use, which tasks will change, what remains their responsibility, and whether AI is designed to help them, or eventually replace them.

The gap between leadership’s perception and employees’ reality adds another layer to her challenge. Executives estimate that only 4% of employees use generative AI for at least 30% of their daily work. Employees put the figure at 13%, more than three times higher. AI adoption may already be moving faster inside organizations than senior leaders can see or govern.

Alexandra must translate broad ambition into daily action, even when the ambition itself is evolving. She has to decide which tools deserve attention, what work can safely be delegated to AI, what should remain firmly human, and how to help people adapt without making them feel diminished or left behind.

That is why Alexandra’s position matters. She is not the enthusiast who sees only possibility, nor the skeptic who sees only risk. She is the person in the middle, trying to make change practical before she has complete answers. Her tipping point will not arrive through one major decision. It will emerge through hundreds of smaller ones: every tool she approves, every process she redesigns, every concern she hears, and every conversation she has with a team looking to her for direction.

As the bus continues along the road, I glance at Alexandra through the rear-view mirror.

“What keeps you up at night?” I ask.

She closes her laptop for a moment.

“I don’t know what I don’t know.”

Bob, the AI Ambassador: “We Are at Risk of Losing Out”

Bob sits at the front, close to me, with a highlighted printout in his lap and several AI newsletters open on his phone. He is the firm’s internal AI strategist, the person expected to define an AI strategy, understand what competitors are doing, monitor the technology landscape, and help the organization decide when and where to invest.

It is not an easy role. AI evolves so quickly that yesterday’s insight can feel incomplete by tomorrow morning. New models, vendors, capabilities, and claims arrive constantly. Bob reads, attends training sessions, speaks with technology providers, and tries to distinguish what is genuinely useful from what is simply generating excitement.

His main concern is not whether AI will matter. For Bob, that question has already been answered. The harder question is how the company can turn AI investment into business value. What is the value of acquiring a powerful tool if the people expected to use it are not prepared? What is the benefit of a sophisticated model if it does not improve decisions, customer service, productivity, quality, or speed? How can the organization make investment decisions today when the technology may look entirely different in six months?

“I need to be ahead of the game,” Bob says. “Otherwise, we are at risk of losing out.”

He does not say it with panic, but there is urgency in his voice. He knows doing nothing may create a competitive disadvantage. But he also knows that doing too much, too quickly, can lead to wasted investments, disconnected tools, duplicated efforts, and people using technology without clear direction or safeguards.

Bob sees the AI tipping point as an organizational challenge more than a technical one. The technology may be extraordinary, but the organization still has to decide how it will work differently. It needs new skills, processes, governance, and ways to measure value. The company cannot simply buy AI and expect transformation to happen on its own.

Jolie, who has been listening quietly from the middle of the bus, leans forward.

“Bob, what’s next?” she asks.

Bob looks out at the road ahead.

“Faster, better, smarter AI,” he says. “The sky is the limit.”

Jolie nods.

“What is going to be a barrier?”

“Costs and privacy,” Bob replies. “The more capable AI becomes, the more expensive it may be to build, operate, govern, and secure. And the more deeply it enters our work, the more carefully we will need to protect information, customers, employees, and the company.”

Jolie considers that before asking one final question.

“Who should we partner with?”

Bob closes the printout in his lap.

“We will need multiple partners,” he says. “AI is not a solution we can simply buy and install. It will force us to change our operating model across all aspects of the company. We will need technology partners, data partners, implementation partners, risk and security partners, and people who can help us build the skills to use AI responsibly. The question is not only who we partner with. It is how we change ourselves.”

Jolie, the CEO: “We Don’t Have the Granularity, Yet”

Jolie sits in the middle of the bus with her phone turned face-down on the seat beside her. For once, she is listening more than speaking. As CEO, people expect her to have clarity, direction, and answers. Yet AI has put her in an unfamiliar position: she sees enormous possibility but is not entirely certain where to begin.

“AI is important,” Jolie says. “I do not think anyone doubts that anymore. But we do not yet have the granularity to understand how tactically important it is for every part of our business. And once we have the technology, what exactly are we going to do with it?”

She describes the feeling as being a child in a candy store. There are countless possibilities: improving customer interactions, accelerating analysis, automating internal processes, supporting employees, strengthening decisions, reducing cost, and creating new products or services. The challenge is not a shortage of ideas. It is deciding which opportunities matter most, which can create value quickly, and which are worth the investment required to pursue them properly.

Jolie knows she is not alone in this uncertainty. Across the corporate world, AI pilots are being launched at remarkable speed, but many struggle to become sustainable, large-scale solutions. Around 95% of generative-AI pilots are estimated to fail to produce meaningful business impact. Only 25% of initiatives achieve the return on investment leaders initially expect. At the same time, 84% of executives say they are seeing some return from AI investments, while only 39% can attribute any impact on enterprise profit to AI so far.

For Jolie, that gap is the central issue. A company may see early productivity benefits, faster outputs, and better access to information. But are the gains large enough to justify the investment? Are they repeatable across the organization? Can they be measured and sustained as the costs of technology, data, security, talent, governance, and infrastructure continue to evolve?

What worries her most is not simply the initial investment. She can build a business case for software, infrastructure, training, and implementation. The harder question is the recurring cost of becoming, and remaining, an AI-enabled company. Will AI be as powerful, accessible, and affordable tomorrow as it is today? Or will the economics shift as demand for computing power, data, specialized talent, and secure systems rises?

Jolie understands that the tipping point is not necessarily a future event. In many ways, it is already here. It is the moment when leaders must move beyond enthusiasm and decide how to translate investment into lasting value. It is when AI stops being a set of interesting tools and becomes an operating-model question: how the company makes decisions, serves customers, organizes work, manages risk, and develops people.

Jim, listening from the window seat, turns toward Jolie.

“What do you think the competition is doing?” he asks.

Jolie looks ahead for a moment before answering.

“Our biggest competition is us at this point,” she says. “This is going to be so big and so fast that we will lose ourselves if we do not strike the right chord.”

Maggie, the AI Expert: “We Need AI, But Not Too Much AI”

Maggie sits near the back, away from the loudest conversations. She has listened as Jim spoke about time, Alexandra explained the uncertainty of managing change, Bob described the pressure to stay ahead, and Jolie questioned how to convert investment into value.

When Maggie finally speaks, the bus becomes quieter.

She works close to the frontier of artificial intelligence. She understands what the technology can do today, what it may be able to do tomorrow, and, just as importantly, what no one can yet fully predict. Her concern is not that AI is inherently bad. She does not argue that organizations should stop using it, nor does she deny its potential to improve work, science, services, and decision-making.

Her concern is more fundamental.

“Something is going to snap,” Maggie says. “We need AI, but not too much AI.”

To Maggie, the risk is not contained in one dramatic scenario. It lies in the accumulation of pressures: technology is improving rapidly; companies are investing enormous sums; governments are trying to write rules after the technology has moved ahead; workers are trying to understand how their roles will change; and societies are questioning who controls the systems, the data, and the benefits.

She describes AI as a race in which the competitors are building the track, setting the rules, selling tickets, and deciding who is allowed to enter, all at once. The speed is extraordinary, but the direction is not always clear.

Maggie points to the scale of the anomaly. In 2024, global data centres consumed approximately 415 terawatt-hours of electricity, or about 1.5% of global electricity consumption. By 2030, that demand could rise to approximately 945 terawatt-hours. At the same time, the largest technology companies are expected to invest roughly $775–800 billion in AI-related capital expenditure in 2026 alone.

Those figures do not prove AI will fail. They show how quickly it has become more than a software story. It is now an energy, infrastructure, investment, workforce, and social story.

The technology requires data centres. Data centres require power. Power requires infrastructure, approvals, investment, land, and sometimes difficult choices about where energy should go. If AI demand rises while countries seek to electrify transport, expand manufacturing, and meet climate commitments, competition for energy and capital will become more visible.

There is also economic tension. Companies are spending heavily because they believe AI will transform their industries. Some investments will create exceptional value. Others will be too early, too expensive, poorly targeted, or impossible to scale. Maggie does not believe every organization needs to be first. But she believes every organization needs to understand what it is building toward.

“The danger is not that people will invest in AI,” she says. “The danger is that they will invest without understanding the full system around it.”

For Maggie, that system includes cybersecurity, privacy, intellectual property, data quality, bias, workforce disruption, misinformation, concentration of economic power, and the possibility that decisions once made by people will increasingly be delegated to systems few can explain. It also includes a deeper question: when does helpful automation become dependence?

She explains that the tipping point may not come because one model suddenly becomes too powerful or one company collapses. It may come through a combination of smaller events: a major privacy breach, a high-profile cyberattack, a sudden increase in energy costs, a public backlash against job displacement, an investor deciding returns are taking too long, or a regulator imposing restrictions that reshape an entire sector.

That is why Maggie prefers the idea of a tipping point to a cliff. A cliff suggests one edge and one fall. A tipping point suggests that the system changes because too many pressures meet at once. It may force the industry to build more efficient models, adopt clearer regulation, slow down certain uses, rethink investment levels, or create new operating models.

Maggie is not pessimistic. She is asking for discipline.

“We need to make AI useful,” she says. “We need to make it safe enough, understandable enough, affordable enough, and sustainable enough that society can live with it. That requires more than better models. It requires better decisions.”

Jolie turns in her seat and looks toward Maggie.

“When will be the first tipping point?” she asks.

Maggie pauses before responding.

“I don’t know,” she says. “But keep an eye on major investors. They will react first. One thing is for sure: there will be a tipping point.”

Diamond, the Futurist: “Two Kinds of Diamonds”

By the time the bus approaches Cotonou, the afternoon light has softened and the conversations have become more reflective. Jim has spoken about time. Alexandra has spoken about uncertainty. Bob has spoken about urgency. Jolie has spoken about investment. Maggie has warned of the tipping point.

Diamond has listened to them all.

He is less concerned with this quarter’s investment plan or the next board presentation. His attention is fixed on a larger question: what will AI do to the next generation of leaders, not only in business, but in schools, families, institutions, and societies?

Diamond does not see AI as only a technical tool. He sees it as a force that could change how people learn, work, value expertise, and define success.

“People keep asking what AI can do,” he says. “I am more interested in what AI will make people become.”

He turns toward the group and introduces the analogy that earned him his name.

“In the market, there are two types of diamonds,” Diamond says. “The first is valuable because it is rare. Its availability is constrained, and that scarcity is part of its value. The second is mass-produced. It may look similar, but it is created through replication, volume, and access. Its value is shaped differently.”

He pauses.

“That is what I think will happen with AI.”

Today, powerful AI still feels exclusive. It requires advanced computing infrastructure, specialized talent, large investments, enormous volumes of data, and access to technology controlled by a relatively small number of organizations. It is moving quickly, but it is not yet equally available to everyone. The most advanced systems remain concentrated among major technology companies, governments, investors, research institutions, and ambitious organizations trying to enter the space.

But Diamond does not believe that will last forever.

“Today, AI is exclusive,” he says. “Tomorrow, it may be everywhere. The question is not whether it becomes more accessible. The question is what happens when it does.”

That is what he calls the “discounted AI” moment: the point at which powerful AI capabilities become affordable, ordinary, widespread, and embedded into almost every part of life. It will no longer be enough for an organization to say it uses AI. That will be like saying it uses electricity, email, or spreadsheets. The differentiator will be how well people use it, how responsibly they govern it, and how effectively they combine it with human judgment.

For Diamond, this is where the real tipping point sits. It is not necessarily the launch of a new model or another company raising billions of dollars. It is the moment when AI becomes so widely available that it changes the behavior of entire markets. When everyone has access to intelligence at scale, what becomes valuable? If analysis can be produced in seconds, what happens to the analyst? If content can be created instantly, what happens to creativity? If every leader can ask an AI system for strategy, what distinguishes an exceptional leader?

The answer, Diamond believes, will not be technical knowledge alone. It will be judgment, curiosity, discipline, ethics, and imagination. It will be the ability to ask better questions, recognize when not to trust an answer, make decisions when data is incomplete, and understand how decisions affect people who are not in the room.

The workforce implications are substantial. By 2030, an estimated 92 million jobs may be displaced, while 170 million new roles may be created, a net gain of 78 million jobs. But it will not feel like a gain to people whose skills, confidence, and security are disrupted before new opportunities arrive. At the same time, employers expect 39% of workers’ core skills to change by the end of the decade.

Diamond believes those figures should shift the discussion away from whether AI will replace people and toward how people must be prepared to work with it.

“Technology does not eliminate the need for people,” he says. “It changes the type of people organizations need.”

He thinks about students using AI before schools have fully decided how to teach with it. He thinks about young professionals entering workplaces where the first draft of almost anything, an email, report, presentation, analysis, or plan, can be produced in moments. He thinks about leaders who will have access to more information than any generation before them, yet may struggle to determine what deserves trust.

The education challenge is not simply to teach young people how to use AI tools. It is to ensure they retain the ability to think without them. They must learn to reason, communicate, collaborate, challenge assumptions, understand context, and take responsibility for decisions. Otherwise, society risks creating a generation that can generate answers quickly but cannot recognize when those answers are incomplete, biased, harmful, or wrong.

Diamond sees the same challenge inside organizations. Companies cannot treat AI as a separate technology agenda run only by IT, digital, or innovation teams. They need to rethink talent, leadership, operating models, training, performance management, decision rights, risk management, and culture.

AI may automate a task, but it cannot define the values behind it. It may recommend an action, but it cannot carry moral responsibility for the consequence. It may make an organization faster, but it cannot decide what that organization should be fast at doing.

As the bus slows and the lights of Cotonou become visible, Jolie turns toward Diamond.

“Are you saying that whatever happens, it is always about people?” she asks.

Diamond smiles.

“Yes,” he says. “It is always about people: how we select them, train them, and treat them. Unless we no longer want to be a part of this world, people will always be the central part of every human adventure.”

He looks around the bus one final time.

“I think every organization has to focus on people with AI, as opposed to AI, AI, and AI.”