Is Sam Altman’s $7 trillion chip dream real or a ruse?

In an attempt to generate between $5 trillion and $7 trillion in capital, OpenAI CEO Sam Altman is now holding conversations with investors, including those from the United Arab Emirates. The objective is to improve AI capabilities and boost global chip production capacity.

Let’s look at what exactly $7 trillion might fund and why he needs it.

Willy Shih, a professor at Harvard Business School and former IBM employee, stated that the AI models underlying ChatGPT do need a lot of processing power—more than most people realize.

According to rumors, part of Altman's plan is to establish a joint venture with power suppliers, investors, and chip makers to fund the establishment of chip foundries, which the chip makers would subsequently run.

$7 trillion is a lot of money, even if Altman is trying to raise money (and maybe debt) by the salami approach, say, by breaking it up into twenty installments of $350 billion spread over ten years. For comparative purposes, this sum exceeds the GDP of both Germany and the United Kingdom, 45 times the cost of constructing the International Space Station, twice the worldwide car industry's 2023 revenue, and 13 times the global semiconductor industry's 2023 revenue.

Altman might spend trillions on data centers, which house the GPUs required to train the AI models that power businesses like ChatGPT and Sora, if his goal is to create larger models for OpenAI. According to research firm IDC, the U.S. market for data center development was estimated to be worth $24.63 billion in 2024. He could thus purchase forty times as many data centers as there are now if he spent $1 trillion on chips.

It seems that, according to Altman, AI is the future of technology and the world itself. And he will stop at nothing to make his dream a reality.