Jeff Bezos has been busy. According to news reports, the founder of Amazon.com Inc. is seeking commitments from sovereign wealth funds and asset managers in the Middle East and Singapore for his proposed US$100 billion industrial transformation fund. Apparently, this will buy manufacturing companies and transform their operations using artificial intelligence.
It’s not clear how receptive global investors are likely to be, given current geopolitical tensions and the potential for more disruption between the US and global trading partners. But for better or worse, there is no doubting the size of the opportunity.
Bridgewater Associates has already forecast total demand for AI-related infrastructure in 2026 at some $650 billion, up from the $410 billion spent in 2025. Set against those figures, Bezos’s $100 billion target suddenly doesn’t seem very excessive.
Furthermore, like cloud computing, AI is becoming a resource and input for many companies, albeit an increasingly expensive one. AI buildout and capacity bottlenecks are helping to drive up prices and intensify competition between customers.
All this should bode well for Bezos’s fund. And yet, in size and focus, and target investor base, it resembles the $100 billion SoftBank Vision Fund, which lost 18% of its value in 2020, and whose successor fund had to be scaled back to less than a third of the first fund.
Moreover, this may not be the best time for seeking further commitments to AI-related propositions.
Valuations are increasingly volatile. SK Hynix, the South Korean semiconductor company, saw its market cap more than double between September 2025 and March 2026 as market assessments of the hardware needs of AI platforms shifted to $461.86 billion versus some $292 billion for China’s Alibaba Group.
Other entities exposed to AI are doing less well. Microsoft Corp. lost $613 billion of its market cap in February 2026 over fears that its AI business was under threat from competitors.
Amid these seesawing valuations, Jim Rickards, an economist and former White House adviser, predicted an AI bubble that will burst by the end of April and trigger an 80% stock market crash. OpenAI’s “code red” warning last December over intense competition from Google’s Gemini 3 may or may not be the last straw.
It’s not clear whether AI will achieve all that its pundits promise, but it is clearly disruptive, and vulnerable to disruption. Valuations and business models may change overnight. The dot-combust in early 2000 led to a market collapse, though leading technology companies worked through this to dominate today’s indexes.
Of course, Bezos’s AI infrastructure fund may come together just in time to buy assets near the bottom of a falling market. But the dangers of overexposure to AI have never been more apparent. And in today’s febrile conditions, Rickards’s prediction may well turn out to be a self-fulfilling prophecy.
