DeepSeek has unravelled the assumption that a few hyperscalers will dominate the artificial intelligence market, extracting payments from an AI-dependent global economy. The Chinese company’s open-source program lets businesses develop their own applications without having to pay anyone. The revenue assumption of hundreds of billions of dollars that underpins hyperscalers is looking like hot air.
Advertisement
The US economy depends on its buoyant stock market which, at twice the value of its gross domestic product, is well above the historical average. Tech stocks are a big part of this lofty valuation, justified by the US lead in large language models (LLMs): AI models that can understand and generate human language. The assumption? That the global economy would become dependent on them and American AI companies would dominate.
A few so-called hyperscalers are still raising hundreds of billions of dollars to build massive data centres and power plants to develop LLMs and lock in their lead with sheer financial power and the US chip advantage. The massive spending is supposed to lead to artificial general intelligence (AGI), aka digital gods.
The talking heads want us to believe that AI will provide answers to everything. Hence, consumers and businesses must ask the AI gods to make their decisions. With so much power, the US can lock in its dominance and extract a big slice of the global economy to pay off its national debt. The extreme US stock market valuation is merely pricing in the coming bonanza.
But it is debatable how far LLMs – so far, glorified search engines – can go. Can an LLM ever take over the physical world? There is no evidence it can. And if it remains a tool for information manipulation, how much could it improve productivity?
Advertisement
Smartphones have led to an astronomical rise in information consumption. But productivity has levelled off around the world. Too much information consumption doesn’t seem good for productivity. Nevertheless, the dream is needed to justify the enthusiasm.