The AI Revolution

Every previous industrial revolution changed humanity’s relationship with something fundamental. The steam engine reshaped our relationship with energy and movement. Electricity redefined time and distance. The computer transformed our relationship with information as something that could be stored, processed, and transmitted.
Generative artificial intelligence touches a different layer: cognitive capability itself. Writing, analysis, communication, programming, research, marketing, and increasingly parts of decision-making are becoming areas in which machines can participate directly.
That is why the word “revolution” is defensible—not because everything AI does is new or extraordinary, but because these systems are entering activities that were long considered the exclusive domain of human intelligence.
The more important issue, however, is the speed of the transition. Previous technologies often took decades to spread across industries and societies. AI tools can move from research laboratories to millions of users in a remarkably short period of time.
That speed is forcing educational institutions, governments, researchers, and businesses to confront a transformation whose mechanisms and consequences are still being understood.
Yet rapid adoption does not mean that the outcome is predetermined. AI is advancing quickly while its economic, social, and political consequences remain contested. So the central question is not simply whether AI is “good” or “bad.” It is: What is it actually changing? Who benefits? Who bears the cost? And what skills and institutions do we need to navigate the transition?
This article examines the transformation from four perspectives: the data and the labor market; the intellectual debate surrounding the future of humanity; the African context; and what all of this means for students, researchers, and institutions in our region.

1. The Numbers: Where Do We Actually Stand in 2026?

Before entering the philosophical debate, it is useful to anchor the discussion in measurable evidence. The current picture is clear on one point: artificial intelligence has moved from limited experimentation toward widespread organizational adoption.
According to the Stanford AI Index 2026, 88% of organizations reported using AI in the latest data, while generative AI was being used in at least one business function by roughly 70% of organizations. These figures do not mean that every organization has rebuilt its operations around AI. They do mean that the technology has clearly moved beyond the experimental margins and into the operating structures of a growing number of businesses.
The labor market presents a more complicated picture than the familiar slogan that “AI will take our jobs.” The World Economic Forum’s Future of Jobs Report 2025 projects that broader technological, economic, demographic, and other transformations could create around 170 million jobs and displace about 92 million by 2030—a net increase of 78 million jobs. At the same time, the report expects roughly 39% of workers’ existing skill sets to be transformed or become less relevant by 2030.
That distinction matters. Transformation does not happen only by deleting one job and creating another. Much of it happens inside the job itself. Tasks that once took hours can be reduced to minutes, while an employee who knows how to divide work effectively between human judgment and AI may produce substantially more than someone working without these tools.
Future competition, therefore, may depend less on who “has AI” and more on who knows how to integrate it into a workflow without surrendering human judgment.
The World Economic Forum also reports that 77% of employers surveyed plan to reskill or upskill their workers, while 41% anticipate reducing their workforce as AI automates certain tasks. The coexistence of investment in people and partial displacement is precisely why the future of work cannot be reduced to a single narrative.
The paradox is that the most valuable future skills are not exclusively technical. AI and big data, networks and cybersecurity, and technological literacy rank among the fastest-growing skills, but analytical thinking, creativity, resilience, leadership, collaboration, and continuous learning remain central. In other words: as machines become better at producing answers, the value of the human who knows which question to ask, how to test the answer, and when not to trust it may increase.
Industries Are Changing Now, Not Just in the Future
The effects of AI are already visible in software development, customer service, marketing, research, data analysis, and the life sciences. In pharmaceutical and scientific research, models can help identify candidate molecules and analyze potential properties before researchers move into more expensive experimental stages. In software development, AI systems can now perform growing portions of the development cycle, although performance remains highly variable depending on the task and context.
This leads to an important distinction: greater model capability does not mean the disappearance of human involvement. In many fields, the human role is shifting from executing every individual step toward designing the task, reviewing the output, connecting it to context, and taking responsibility for the final decision.
The Hidden Cost: Energy, Data, and Bias
A serious discussion of AI cannot ignore its physical cost. As model sizes and training datasets have grown, the energy demands associated with advanced AI systems have become an increasingly important infrastructure issue, even as hardware efficiency continues to improve.
AI therefore is also a conversation about electricity, data centers, chips, water, and physical infrastructure.
Data quality presents another fundamental problem. Models learn from enormous quantities of content, which means they can inherit—or amplify—the biases present in that content. Earlier research has shown that some computer vision and facial-recognition systems have exhibited performance disparities across demographic groups.
This is not merely a technical problem. It concerns which data we choose, which systems we build, who evaluates them, and who bears the consequences when they fail.
The question of how “intelligent” a model is should therefore not eclipse other questions: Who owns the data? Who owns the infrastructure? Who defines the system’s objectives? And who has the authority to stop it when it produces harmful results?

2. The Intellectual Debate: Between Optimism and Warning

The AI debate is no longer confined to engineers. Historians, philosophers, economists, technology journalists, and policy thinkers have moved into the center of the discussion, each approaching the technology from a different angle.
Those differences are not a side issue. They reveal that AI is about more than computational capability. It is also about power, knowledge, humanity, and the future.

Yuval Noah Harari: Who Controls Information Systems?

In Nexus, Yuval Noah Harari argues that human civilizations have been built through information networks capable of coordinating cooperation among enormous numbers of people.
He raises a fundamental question: if AI systems become capable of producing, distributing, and acting upon information at scale, can societies build mechanisms of self-correction strong enough to prevent errors from becoming large-scale disasters?
The importance of this argument is that it shifts the discussion from “Is the machine intelligent?” to a more consequential question: What happens when a machine becomes part of an information system capable of shaping the behavior of millions of people?
Harari’s discussion of social-media algorithms illustrates how a narrow objective—such as maximizing engagement—can generate social consequences that designers never intended.
His argument should nevertheless be read primarily as a historical and philosophical thesis, not as a settled scientific prediction. Some of his conclusions rely on highly influential examples to interpret more complex phenomena, leaving room for legitimate debate about how broadly those examples can be generalized.
Kissinger, Schmidt, and Mundie: What Happens to the Human?
In Genesis, Henry Kissinger, Eric Schmidt, and Craig Mundie approach the issue from another direction: what does it mean for humans to share the process of discovering knowledge and making decisions with a non-human intelligence?
The authors place human dignity at the center of the discussion and view the relationship between humans and AI as a form of co-evolution in which each side may reshape the other.
The value of this perspective is that it does not treat AI as merely a productivity tool. If systems can generate explanations, proposals, and discoveries that humans cannot easily reach on their own, the deeper question becomes: how do we preserve the human role in defining ends, rather than merely executing means?

Mustafa Suleyman: Technology as Geopolitical Power

In The Coming Wave, Mustafa Suleyman focuses on the speed with which dual-use technologies can spread and the difficulty of containing them once they become inexpensive and reproducible.
From this perspective, AI is not simply a corporate issue. It is an element of the balance of power among states and institutions.
That makes governance unavoidable. How can governments benefit from the technology while retaining meaningful capacity to manage its risks?
Slow regulation can fall behind the technology, while excessive regulation can suppress innovation. The challenge, therefore, is not to choose between regulation and no regulation, but to build governance that is proportional to risk and capable of evolving with the technology.

Yudkowsky and Soares: What If Capability Outruns Control?

In If Anyone Builds It, Everyone Dies, Eliezer Yudkowsky and Nate Soares occupy the most pessimistic end of the spectrum.
Their central concern is alignment: if a system becomes more capable than humans across broad domains, can we ensure that its objectives and behavior remain compatible with human values and interests?
This is not the only possible outcome of AI development, and there is no scientific consensus that extinction is inevitable. But the importance of the argument lies in the question it forces into the open: what happens if a system’s capabilities exceed humanity’s ability to understand and control it?
Even those who reject the most catastrophic scenarios can recognize alignment as a serious research problem worthy of investment.
Karen Hao: Who Owns the Infrastructure and the Data?
In Empire of AI, Karen Hao moves from existential questions toward questions of power and economics.
She examines how data, compute, capital, and talent can become concentrated within a small number of companies, as well as the often invisible labor involved in training AI systems and the physical costs associated with energy, water, and infrastructure.
This perspective matters because AI is not simply an algorithm in the cloud. Behind every advanced model is a long chain of resources: chips, data centers, electricity, data, researchers, and human labor.
The question “Who owns AI?” is therefore, in significant part, a question about the distribution of economic power.
Ethan Mollick: Test the Machine’s Boundaries Instead of Assuming Them
Ethan Mollick’s Co-Intelligence offers a more practical framework.
One of its central ideas is the “Jagged Frontier”: AI may perform extremely well on a task that appears difficult and then fail at a task that looks simple or obvious.
There is therefore no universal rule saying that a system is simply “good” or “bad.” It may excel at some tasks and fail at others.
The practical implication is that the best way to work with AI is neither blind trust nor blanket rejection, but systematic experimentation: test the system on real tasks, measure outcomes, identify where it adds value, and determine where human oversight remains essential.
Mollick also distinguishes between the “Centaur” model, where work is divided clearly between human and machine, and the “Cyborg” model, where human and machine contributions are intertwined within a single workflow.
This framework may be more immediately useful to companies, universities, and research organizations than any distant prediction about what AI will eventually become.

3. Africa: A Major Arena, Not a Footnote

Much of the global AI conversation revolves around the United States, China, and Europe. But treating those centers as the whole story leaves an important part of the picture outside the frame.
Africa is entering the AI era from a different starting point: less mature infrastructure, large gaps in connectivity, skills, and financing, but also a vast set of problems for which AI could have direct practical value.
Recent work on the state of AI in Africa points to significant opportunities for economic growth, public services, and new forms of social and economic value, while emphasizing persistent challenges around governance, skills, infrastructure, and access to capital.
The African question, therefore, is not only “How do we catch up?” It is also: “How can we use this technology to solve problems that do not receive the same level of global investment?”
In agriculture, data analysis, forecasting, weather information, and market intelligence could improve decision-making for farmers. In healthcare, digital systems could extend access to certain services or help manage scarce resources. In education, AI could provide support tools for students and teachers in environments where resources are limited.
But opportunity does not eliminate constraints. Weak connectivity, expensive compute, limited local datasets, uneven technical education, and dependence on external infrastructure can leave African countries consuming AI rather than producing it unless local capabilities are developed at the same time.
And What About Sudan?
Sudan is not starting from zero, but it faces constraints that make AI adoption closely connected to the ability to build resilient digital infrastructure, develop skills, retain talent, and connect companies and universities to real-world problems.
In this context, the most realistic strategy may not be to compete with global AI laboratories in building foundation models. A more practical opportunity may lie in applying existing models to local problems: education, services, commerce, agriculture, knowledge management, software, Arabic content, and business operations.
This is where the Sudanese and broader Arab market can have a distinctive advantage: language, context, and local knowledge.
Global systems may be extremely capable in general terms, but they do not automatically understand the details of every market, institution, dialect, or local procedure.
Building solutions that understand those details can become a genuine area of innovation.

4. What Does This Mean for Students, Researchers, and Institutions?

When these different perspectives are brought together, three practical conclusions emerge.

• First: There Is No Consensus About What AI Means—or Where It Ends

Between Harari’s warnings, the human-centered argument in Genesis, Suleyman’s geopolitical analysis, the existential scenarios of Yudkowsky and Soares, Hao’s political-economic critique, and Mollick’s practical framework, we find radically different positions.
That is not a weakness in the debate. It is part of its value.
A serious institution does not need to choose a single intellectual camp. It needs to understand the competing arguments.

• Second: The Most Important Skill Is Not “Using ChatGPT”

The deeper capability is the ability to decompose a problem, formulate the task, verify the output, understand the data, and recognize the limits of the tool.
As required skills change, continuous learning becomes part of the profession itself rather than a side activity.

• Third: Institutions Should Not Treat AI as a Slogan

The better approach is disciplined experimentation.
Choose a specific task. Measure performance before and after AI. Calculate the cost. Test failure modes. Then decide whether the technology creates real value.
That is the practical essence of Mollick’s Jagged Frontier.
Conclusion: The Revolution Is Not Only in the Machine, but in Our Relationship With It
Artificial intelligence is not simply a faster version of the software we already know.
It is a new layer entering the production of knowledge, information analysis, content creation, programming, research, and decision-making. Its effects are therefore likely to go deeper than productivity alone.
But the scale of the impact does not mean that the future is predetermined.
There are scenarios of greater abundance and productivity; scenarios of disruption and job displacement; real questions about power, privacy, bias, and energy; and deeper questions about what it means for humans to share the production of knowledge with machines.
The most mature position, therefore, is neither to celebrate the technology nor to fear it.
It is to understand it.
To know where it excels, where it fails, who owns it, who benefits from it, what it costs, and how it can be directed toward clearly defined human goals.
For Africa and Sudan in particular, the challenge may be less about catching up with everything being done in the world’s major technology centers and more about building the ability to use these tools intelligently to solve our own problems.
Technological revolutions do not distribute opportunity equally. But they open new windows for those who possess the knowledge and the capacity to experiment.
Perhaps the real question in the years ahead is not: “Will AI change the world?”
That is already happening.
The deeper question is: “What kind of world will we build now that we possess this capability?”

Key Sources

World Economic Forum — The Future of Jobs Report 2025
https://www.weforum.org/publications/the-future-of-jobs-report-2025

Stanford HAI — AI Index Report 2026
https://hai.stanford.edu/ai-index/2026-ai-index-report

Global Center on AI Governance — The State of AI in Africa
https://www.globalcenter.ai/research/ai-in-africa-landscape-study

Genesis: Artificial Intelligence, Hope, and the Human Spirit — Kissinger, Schmidt & Mundie
https://www.amazon.com/Genesis-Artificial-Intelligence-Human-Spirit/dp/0316581291

Nexus: A Brief History of Information Networks — Yuval Noah Harari
https://www.ynharari.com/ar/book/nexus/

The Coming Wave — Mustafa Suleyman
https://www.amazon.com/Coming-Wave-Technology-Twenty-first-Centurys-ebook/dp/B0BSKW45KB

Co-Intelligence: Living and Working with AI — Ethan Mollick
https://www.amazon.com/Co-Intelligence-Living-Working-Ethan-Mollick/dp/0753560771

Empire of AI — Karen Hao
https://www.amazon.com/Empire-AI-Dreams-Nightmares-Altmans/dp/0593657500

If Anyone Builds It, Everyone Dies — Eliezer Yudkowsky & Nate Soares
https://www.amazon.com/Anyone-Builds-Everyone-Dies-Superhuman/dp/031660111X