The tech giant is all-in with artificial intelligence (AI). Microsoft (msft -1.90%)) Recently, we announced plans to spend $80 billion this year on building AI data centers to support advanced AI models in training and deploy those models to the cloud. Metaplatform follows the lawsuit, with up to $65 billion allocated to AI infrastructure in 2025.
To avoid losing, Amazon has raised its annual capital spending plan to $100 billion, but not all of this is directed towards AI computing capabilities. Finally, Alphabet raised its capital expenditure plan to $75 billion in 2025. Many of them were related to AI.
In addition to all this spending, the Trump administration’s Stargate project will involve an additional $500 billion in AI investment over the next four years, with the European Union mobilizing 200 billion euros ($209 billion) for AI investments. We are aiming for this. Simply put, a huge amount of capital is rapidly spent on AI data centers and the equipment needed to train and run powerful AI models, especially Nvidia’s Graphic Processing Units (GPUs). (NVDA) -4.05%)).
Raining at Nvidia parade
Demand for Nvidia’s data center GPUs could remain strong for the time being as businesses and governments become more intense in AI spending. What happens after that? It depends on whether demand for AI computing capacity rises to meet an explosive amount of supply.
In a recent interview, Microsoft CEO Satya Nadella said she was curious to see Nvidia investors have to pause. While Microsoft has invested heavily in AI computing capabilities, Nadella essentially predicted an oversupply over the next few years. Rather than building its own additional capabilities, the company plans to lease “many capacity” in 2027 and 2028. This is the kicker:
“Because I look at the build and say, ‘This is great.’ The only thing that happens with every calculated build is that the price drops. ”
One way bubbles can form is through overinvestment. Businesses and governments are building massive AI computing capabilities independently without a real idea of what demand will look like in a few years. Once this current buildout is complete, supply requires little miscalculation to significantly outweigh demand.
The closest historical similarity to this AI-based Bonanza is not the dot com bubble, but the railway bubble from the late 1800s. Too much capital spent building rail capacity led to a huge oversupply that helped send the rail to bankruptcy and spark financial panic. Railways, like AI, were innovative technologies that fundamentally changed the world, but that fact did not save investors.
In an oversupply scenario, AI computing prices will plummet. This is great for those who lend AI computing capabilities and bad for those who own them. In theory, cheaper access to AI computing could increase demand. However, this effect has limitations. Oversupply of railways did not lead to people who would travel more or ship more freight in order to perfectly match that supply. After all, trees do not grow into the sky.
In the case of Nvidia, the oversupply of AI is double spraying. Not only will pricing for powerful data center GPUs be under pressure, but demand will also fall off the cliff. When AI computing capacity is too high, demand comes primarily from maintaining existing infrastructure rather than expanding. It also has the complexity of increasing numbers of high-tech giants designing their own AI chips. Openai is the latest to reportedly jump into Bandwagon as it aims to reduce its dependence on Nvidia.
Incredibly dangerous inventory
The trillion dollar question is whether demand for AI computing power could rise sufficiently to absorb the vast supply that will become online in the coming years. The fate of Nvidia stocks depends on balance.
Microsoft has taken a sensible approach to AI investment. The company is still building up enough capacity, but is careful not to overdo it. By planning to lease many capabilities in the future, Microsoft is taking considerable risk from the table.
If Nadella is right about oversupplying AI, it’s difficult to see how Nvidia’s revenue and profits can grow at a pace that is sufficient to justify multiple dollar valuations. As AI investment explodes, it may seem brighter in the short term for GPU leaders, but the long term picture is less clear. If we are truly witnessing a large AI overinvestment cycle, Nvidia stocks will sooner or later be subject to serious pressure.
John Mackey, former CEO of Amazon subsidiary Whole Foods Market, is a member of Motley Fool’s board of directors. Randi Zuckerberg, a former director of market development, Facebook spokeswoman and sister to Metaplatform CEO Mark Zuckerberg, is a member of Motley Fool’s board of directors. Suzanne Frey, an executive at Alphabet, is a member of the board of directors of Motley Fool. Timothy Green has no position in any of the stocks mentioned. Motley Fool has positions for Alphabet, Amazon, Meta Platforms, Microsoft, and Nvidia, and is recommended. Motley Fool recommends the following options: A $395 phone at Microsoft for January 2026 length and a $405 phone to Microsoft for January 2026 short term. Motley Fools have a disclosure policy.
