In all Technological revolutionVertigo is often confused with euphoria. It happened with railways in the 19th century, with dot-coms in the 1990s, and now with artificial intelligence (AI), it’s no different. but Fundamental difference It is that this time we are not inflating expectations, but rather building the foundations of a cognitive revolution that was unimaginable. What some call itbubble“It is, strictly speaking, a massive investment in the new productive class of the 21st century.
The numbers say it without metaphors. Adults Technological Heading near 400 billion US dollars Annually – 70% of operating cash flow – to chips, data centers and energy. It’s money in silicon and cement, not in logos. It is pure capital investment in a race to build digital factories that will preserve the environment Overall productivity For the coming decades.
But unlike a bubble Internet From the past, AI is already making a difference. Subscription forms – e.g Co-pilot Or ChatGPT – already generates over $9 billion per year, and if it reaches adoption levels similar to Netflix or Amazon Prime, the potential exceeds $100 billion. actually Enterprise APIsAs companies buy Saas information, they transfer another $8 billion today and are expected to reach half a trillion dollars dollars annually.
Autonomous agents, which are able to carry out tasks and transactions without the need for humans, expect to spend another $150 billion Potential income. Smart advertising, an evolution of the classic search engine, adds another $100 billion to the board. Overall, A economy AI will generate more than $800 billion annually in recurring revenue in the coming years.
And the acceleration in adopting it model Productivity has already been observed in code plugin pilots that increase developer productivity by up to 50%, which translates into added value. 1 billion dollars annually. Already in biotechnology, systems like AlphaFold have been able to map 200 million proteins in two years, accelerating the pace Discoveries Which previously took decades.
For their part, we see how big tech companies grow with stable, not bloated, workforces, where productivity has replaced employment as the driver of expansion. In terms of MacroeconomicsIt is estimated that artificial intelligence will contribute 1.5 points of additional growth in… Global GDP Annually until 2035, about 2 trillion US dollars annually.
Bubble destroys productivity; This cycle doubles it. Every profound transformation begins with a seemingly irrational spending phase. the electrification It burned through capital for 20 years at the beginning of the last century before it paid off. Internet It took a decade to justify its valuation.
Artificial intelligence in it.”J curve“, where the return has not yet been reflected, but infrastructure is already being built at US$400 billion annually, accumulating US$1.2 trillion in the next decade, with the US leading its investments with a ratio of 12:1 compared to China. For this infrastructure to be profitable, it must generate US$600,000 million in revenuewhich is a completely reasonable number compared to the cloud computing business, which 15 years ago seemed like a surplus.
But we must understand that the bottleneck is not in development AlgorithmBut in the energy space, where demand for data centers has already increased by 40%, it could add an additional 70 gigawatts in the US alone. By 2030, the equivalent of 60 nuclear power plants or 200 million cubic meters/day of natural gas. AI needs continuous power, not intermittent. This is why the atom returns to its center blackboard With the United States, Japan, the United Kingdom and even Argentina reactivating their nuclear programs. We must always remember that artificial intelligence is like Steel or railwayhas a material basis. There is no program without electrons.
In short, unlike the past, the current boom is supported by profitability and tangible assets, 90% of which are capex In AI it comes from companies with positive cash flow, not from startups with leverage. It is clear that AI adoption is still in its infancy: only 5% of users pay for AI and less than 10% of users pay for AI. Companies Fully integrated, but its adoption is steadily increasing.
In short, there is no excess supply, and there is a shortage of superchips, data centers and atoms to utilize production capacity. He asks Explosive for everyday use. Therefore, we must understand that artificial intelligence is not the destination of this productivity revolution, but rather the infrastructure that accelerates it. Just as the combustion engine transformed energy into movement at the beginning of the twentieth century, artificial intelligence transforms information into productivity, creating a cognitive revolution.
About The Author
