SerpApi.Org
Key Points
Hyperscalers are rushing to become the leaders of artificial intelligence as agentic AI and physical AI emerge.
Bill Ackman believes that lower token costs are around the corner, which can boost profit margins.
Lower token costs can also increase the demand for AI models, cloud platforms, and other AI resources, translating into higher revenue numbers.
Hyperscalers are approaching $700 billion in AI spending in 2026, which can continue to increase in the years ahead. While some investors are concerned about how long it will take for these investments to yield profits, Bill Ackman threw his hat into the ring as a tech bull.
An X post from the billionaire investor detailed a bullish thesis that assumes token costs will become more affordable. If that happens, it can further justify high capital expenditures and translate into higher profits for tech giants.
Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »
Why token costs matter
AI models like ChatGPT read text prompts as tokens. These tokens are also generated when users ask AI models to create images, videos, and other media.
Each token requires some electricity to be produced. A cloud company like Amazon will buy Nvidia chips that can produce the tokens. Other AI chips produce tokens as well, but Nvidia has the largest market share.
Investors closely follow how many tokens can be generated per watt of electricity consumed to determine how costs may go down in the future. If the amount of tokens generated per watt of electricity consumed doubles, then a hyperscaler's token-related costs are cut in half.
Ackman is waiting patiently for tech giants to optimize their token costs. The early stages of the AI boom have been about maximizing the number of processed tokens, regardless of how many tokens were being produced for each watt of electricity consumed. He believes markets are underestimating this opportunity, and he has been using the disconnect to build positions in hyperscalers.
Hyperscalers can end up with higher margins and better revenue growth rates
In theory, token optimization would let hyperscalers operate with fewer Nvidia chips while serving the same number of customers. However, token optimization won't just improve margins. It will also boost demand for hyperscalers' AI models and cloud platforms.
Ironically, lower token costs can actually increase the demand for AI chips. The Jevons Paradox finds that increased efficiency can translate into more customers, so investors shouldn't expect AI chipmakers to crash amid lower token costs.
The report Ackman shared went on to say that Microsoft processed more than 100 trillion tokens in a single quarter in 2025, which was a 5x year-over-year improvement. That included 50 trillion tokens in one month alone, showing meaningful month-over-month growth. Alphabet also notified investors that it was processing more than 16 billion tokens per minute in Q1.
Tech giants get more out of their existing AI chips, which will be important as agentic AI and physical AI continue to scale. The $700 billion in capital expenditures is a big number, but Ackman is focused on the potential returns from those investments. He sees the light at the end of the tunnel that can yield higher profits and accelerated revenue growth rates.
Don’t miss this second chance at a potentially lucrative opportunity
Ever feel like you missed the boat in buying the most successful stocks? Then you’ll want to hear this.
On rare occasions, our expert team of analysts issues a “Double Down” stock recommendation for companies that they think are about to pop. If you’re worried you’ve already missed your chance to invest, now is the best time to buy before it’s too late. And the numbers speak for themselves:
- Nvidia: if you invested $1,000 when we doubled down in 2009, you’d have $526,675!*
- Apple: if you invested $1,000 when we doubled down in 2008, you’d have $63,609!*
- Netflix: if you invested $1,000 when we doubled down in 2004, you’d have $371,842!*
Right now, we’re issuing “Double Down” alerts for three incredible companies, available when you join Stock Advisor, and there may not be another chance like this anytime soon.
*Stock Advisor returns as of July 20, 2026.
Marc Guberti has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, Microsoft, and Nvidia. The Motley Fool has a disclosure policy.
