Infinite
AI’s big players are spending billions on each other’s services. But even if the circular nature of funding seems opaque, it’s been used before to build growth in other sectors.
Is the AI economy in a circular loop? By now you’ve either seen the memes or attempted to decipher this slightly more complex chart from Bloomberg. Either way, you might be wondering, is any of this real money?
The short answer: yes. But somewhere between a flywheel and a financial hall of mirrors, OpenAI, Nvidia ($NVDA), Oracle ($ORCL) and AMD ($AMD) are pumping billions into each other’s products in a way that makes AI demand feel… well, AI generated.
Let’s break it down:
OpenAI needs GPUs to train models. It buys from NVIDIA, and now increasingly AMD.
To power those chips, OpenAI partners with Oracle, which builds out massive GPU data centers – buying even more hardware from NVIDIA and AMD.
Meanwhile, OpenAI’s largest backer Microsoft ($MSFT) buys chips from NVIDIA and AMD, hosts OpenAI’s models on Azure, and positions itself as a platform for both companies.
Behind the scenes, Broadcom ($AVGO) supplies the high-speed networking and custom silicon that ties these GPU clusters together – a critical but less visible layer of the AI stack.
All companies then report massive growth in AI-related spending. Investors cheer. Stocks rise. Valuations soar.
That’s a simplified version. It doesn’t even include intermediaries like GPU cloud specialist CoreWeave ($CRWV), who also benefit from the infinity loop. The demand engine is self-reinforcing, but that doesn’t mean it's not legit. In fact, it’s been done before.
In aviation’s early days, Boeing ($BA) built aeroplanes, while also helping finance its customers and investing in its suppliers.
Defence firm Lockheed Martin ($LMT) followed the same playbook – integrating with suppliers, sometimes funding smaller firms that made parts or materials it needed.
As Morgan Stanley’s Todd Castagno told Sherwood, ‘It’s an ecosystem that probably does need to self-fund… in order to scale and compete.’ His main concern? These complex transactions make it harder to get a clear view of AI demand.
There is one thing that’s crystal clear: we’re not just watching AI evolve – we’re watching the economy restructure itself around it. For investors, the trick is knowing how to weed out companies capitalising on the hype by adding ‘AI’ to their name without any tangible earnings.
As for the firms driving the loop? They’ve built a system that works for them. And as long as the machine keeps buzzing, there’s no reason it can’t work for you too.
This is not financial advice nor a recommendation to invest in the securities listed. The information presented is intended to be of a factual nature only. Past performance is not a reliable indicator of future performance. As always, do your own research and consider seeking financial, legal and taxation advice before investing.