Over the past few years, the world’s largest technology companies have spent hundreds of billions of dollars building the foundations of artificial intelligence. As explored in the previous chapter, AI Computing Power: Why It Shapes the Future of AI, advanced semiconductor chips have become strategic assets, vast data centres continue to expand across continents, and computing power has emerged as one of the most valuable resources in the technology industry. But none of that capacity is being built for its own sake — it exists to reach a far more ambitious destination: Artificial General Intelligence.
Those investments raise a more important question.
What exactly are they trying to build?
For many of the organisations leading today’s AI race, the answer extends far beyond creating a better chatbot, a faster search engine or a more capable coding assistant. Their long-term ambition is far more significant: Artificial General Intelligence (AGI).
Unlike the AI systems available today, AGI is envisioned as a machine capable of learning, reasoning and solving problems across a wide range of domains without being designed specifically for each individual task. Whether that goal can actually be achieved remains one of the biggest unanswered questions in computer science. Even so, the pursuit of AGI has already become the defining ambition of the modern AI industry.
This marks a new chapter in the story of artificial intelligence.
In the previous articles of this series, we explored how AI works, who controls it, why it became a geopolitical competition, how Taiwan’s semiconductor industry became indispensable, and why AI infrastructure and computing power now determine technological leadership. Together, those chapters revealed how the AI revolution is being built.
This article asks a different question.
Why are the world’s most powerful technology companies investing extraordinary resources into a technology that does not yet exist?
The Goal Behind Every Billion-Dollar Investment
Every major technology company involved in artificial intelligence has its own products, research priorities and business strategy. Some focus on consumer applications, others on enterprise software, cloud computing or scientific research.
Yet behind those differences lies a remarkably similar long-term objective.
OpenAI, Google DeepMind, Anthropic and xAI have all publicly described Artificial General Intelligence as part of their broader vision for the future of AI. While each organisation approaches the challenge differently, all of them are attempting to build systems that move beyond today’s specialised models towards AI capable of handling a much wider range of intellectual tasks.
That shared ambition explains why competition inside the AI industry has intensified so rapidly.
The race is no longer about releasing the next chatbot before a competitor. It is about reaching a technological milestone that many researchers believe could reshape software, science and the global economy.
This also explains why investment continues to grow despite enormous costs.
Training increasingly capable AI systems requires specialised hardware, highly skilled researchers and computing resources operating at an unprecedented scale. Companies continue making those investments because they believe that future breakthroughs could fundamentally change the relationship between humans and machines.
Whether they are correct remains uncertain.
But few of the organisations leading AI research appear willing to risk being left behind.
One Destination, Different Visions
Although the largest AI companies often discuss AGI, they do not always describe it in the same way.
OpenAI presents its long-term mission as ensuring that Artificial General Intelligence benefits all of humanity. The organisation argues that increasingly capable AI systems should ultimately create broad societal benefits rather than serving only a small number of companies or governments.
Google DeepMind approaches AGI primarily as a scientific challenge. Its research focuses on developing increasingly capable AI while emphasising responsible development and long-term safety. For DeepMind, AGI represents both a technological objective and a scientific frontier.
Anthropic places greater emphasis on reliability and alignment. The company argues that creating more capable AI should go hand in hand with making those systems more understandable, predictable and controllable. From this perspective, capability alone is not enough. Advanced AI must also remain trustworthy.
Elon Musk’s xAI has adopted a different emphasis, describing its broader mission as accelerating scientific discovery and advancing humanity’s understanding of the universe. Although its public messaging differs from its competitors, AGI remains central to its long-term ambitions.
These differences reveal something important.
The competition is not simply about building the most powerful AI.
It is also about defining what successful AGI should become.
The Race Towards an Unfinished Destination
Despite years of research and enormous financial investment, there is still no universally accepted definition of Artificial General Intelligence.
Researchers broadly agree that AGI should demonstrate capabilities extending beyond today’s specialised AI systems, but they disagree on where that threshold actually lies.
Some believe AGI should perform most economically valuable intellectual work at or above the level of skilled humans.
Others argue that genuine general intelligence requires much more than strong performance across benchmarks. They believe AGI must demonstrate adaptability, independent reasoning and the ability to solve unfamiliar problems without relying heavily on prior examples.
A growing number of researchers question whether intelligence can even be measured using a single definition.
This uncertainty has created one of the most unusual technology races in history.
Companies are competing to reach a destination that nobody has fully defined.
Unlike a sporting event, there is no finish line waiting ahead.
No international organisation will announce that AGI has officially arrived.
Instead, recognition is likely to emerge gradually through scientific consensus, practical capability and real-world performance.
That uncertainty has not discouraged investment.
If anything, it has encouraged companies to move even faster.
The possibility of reaching such a breakthrough before competitors has become one of the strongest incentives driving modern AI research.
