8 September 2025

Artificial General Intelligence (AGI) refers to machines that can reason, learn, and adapt across domains as flexibly as humans. Unlike today’s narrow AI (systems built for a single task) AGI promises universal capability. Some see it as the next industrial revolution; others caution that it remains speculative at best.

Among those urging realism is Michiel Bakker, a Dutch AI researcher teaching at MIT in the United States. In the Netherlands, he also contributes to the AI Deltaplan, a nationwide strategy to strengthen the country’s position in responsible AI development. In a recent interview with Dutch paper De Volkskrant, Bakker argued there is still “no clear path to AGI”—a reminder that, while headlines proclaim breakthroughs, the reality remains more complex. His voice is part of a global conversation that is shaping research priorities, geopolitical competition, and the way businesses prepare for the future.

What Exactly Is AGI?

AGI—or “strong AI”—is intelligence that can generalize across domains. A narrow AI can master Go, drive a car, or translate languages—but ask it to learn an entirely new skill, and it breaks down. AGI, in theory, could do all of these things, adapting quickly to unfamiliar challenges with human-like reasoning.

The idea goes back to Alan Turing, who speculated about machines that could “think.” But turning that vision into practice is elusive. Current AI excels by pattern-matching at scale: a chatbot predicts likely words, a vision system recognizes pixels, a recommendation engine guesses preferences. None of these systems possess true understanding.

Expert predictions about AGI reflect uncertainty. Surveys suggest some researchers expect progress within the next 10–20 years, while others doubt it will ever be achieved. Defining what “general” intelligence actually means—and how we would recognize it—remains unsettled.

Early Sparks and Real-World Signposts

Even if true AGI doesn’t exist yet, today’s systems show flashes of generalization.

• Large Language Models (LLMs): GPT-4 and its successors can perform well across diverse domains—passing bar exams, generating code, analyzing legal arguments, and even making basic medical diagnoses. Some describe this as “proto-AGI,” though critics point out that LLMs still lack reasoning and common sense.
• DeepMind’s innovations: Projects like RoboCat, which teaches itself new robotic tasks, and AlphaFold, which cracked protein folding, demonstrate growing adaptability and creativity in narrow domains.
• Corporate ambition: OpenAI frames its entire mission around achieving AGI, while Chinese giants like Alibaba and Baidu are investing billions in systems designed to reason more broadly.

For enthusiasts, these examples suggest AGI is “emerging in slow motion.” For sceptics, they underline just how far we are from systems that can truly understand and adapt like humans.

The Global Race

The pursuit of AGI is not just technical—it’s also geopolitical.

• United States vs. China: The U.S. is home to most frontier labs, including OpenAI, Anthropic, and DeepMind. Its strategy is to push for breakthroughs that could leapfrog incremental progress. China, meanwhile, emphasizes practical AI applications, rolling them out across healthcare, commerce, and governance. If AGI takes longer than expected, China’s pragmatic focus could prove more advantageous.
• Europe’s position: The EU’s approach is less about racing and more about regulation. The AI Act, which entered into force in 2024, aims to ensure that all AI—whether narrow or general—aligns with European values of transparency, accountability, and human oversight. Initiatives like the Dutch AI Deltaplan seek to connect research, business, and public institutions to build responsible innovation at scale.

At the same time, global governance debates are intensifying. OpenAI’s Sam Altman has floated the idea of a global regulatory body akin to the IAEA for nuclear power, designed to monitor AGI development. Others propose coalitions like the Multinational AGI Consortium (MAGIC), which would centralize research under strict oversight to reduce risks.

Potential, Pitfalls, and What’s Next

Potential

• Economic transformation: AGI could automate a huge portion of knowledge work, reshaping labor markets and productivity.
• Scientific discovery: A machine capable of generating and testing hypotheses across disciplines could accelerate breakthroughs in medicine, energy, and climate solutions.
• Everyday impact: Imagine assistants that not only answer emails but negotiate contracts, manage projects, or plan complex trips.

Pitfalls

• Overhype: If AGI is treated as inevitable and imminent, businesses and governments may misallocate resources, neglecting the tangible value of today’s narrower systems.
• Trust erosion: Each time “AGI is here” headlines prove misleading, public confidence in AI as a whole suffers.
• Safety and ethics: If AGI capabilities emerge suddenly, without proper safeguards, the risks could be profound—from misinformation to economic shocks.

What’s Next

A growing number of researchers advocate for focusing on hybrid intelligence: combining AI’s strengths with human oversight, rather than chasing full automation. Hybrid systems are already transforming fields like radiology, where AI identifies patterns in scans and humans provide judgment. This direction reflects the popular pragmatic perspective: rather than wait for an elusive breakthrough, build tools that responsibly extend human capacity today.

Even if AGI is decades away—or may never be achieved—the idea itself shapes decisions today. AGI is both a dream and a moving target. For some, it’s inevitable; for others, it’s unlikely. What is certain is that the pursuit of AGI is already reshaping global investment, governance, and imagination.

🔮 Futuristic Use Cases for AGI

For now, let’s just imagine where Artificial General Intelligence could take us. Some scenarios are visionary, others cautionary:

🌍 Global Problem-Solving

• Climate modeling and geoengineering solutions.
• Pandemic prediction and vaccine design within days.
• Global supply chain and resource optimization.

🧠 Science & Knowledge

• Autonomous researcher generating and testing hypotheses.
• Cracking unsolved problems in math and physics.
• Perfect universal translation—even across species.

🚀 Societal & Economic Shifts

• A post-work economy with AGI handling most knowledge labor.
• Governments stress-testing policy outcomes instantly.
• AGI-founded companies running at superhuman speed.

🏥 Human Health & Enhancement

• Personal health concierges predicting illness early.
• Cognitive augmentation as a “thinking partner.”
• Radical longevity through breakthroughs in aging research.

⚠️ Cautionary Tales

• Runaway autonomy—an AGI misinterpreting its goal.
• Information dominance that outpaces human fact-checking.
• Extreme inequality if AGI power is concentrated.

These visions highlight why AGI inspires excitement and caution: it could be humanity’s greatest tool—or its biggest challenge.