Motley Fool Money - Episode Summary: "The Hyperscalers Are Hyper-Spending"

Release Date: January 30, 2025
Hosts: Dylan Lewis, Ricky Mulvey, and Mary Long
Guest: Asit Sharma

Introduction

In the January 30, 2025 episode of Motley Fool Money, host Mary Long engages in an insightful discussion with investment analyst Asit Sharma about the substantial capital expenditures by major tech giants, particularly Meta and Microsoft, focusing on their aggressive investments in artificial intelligence (AI) infrastructure. The episode delves into the implications of these expenditures, the emergence of new players like Deep Seek, and the future prospects of AI-driven technologies such as AI assistants and smart glasses.

Meta's Aggressive Investment in AI Infrastructure

Mary Long kicks off the discussion by highlighting Meta's significant increase in capital expenditures aimed at bolstering its AI capabilities. Meta has announced plans to invest between $60 to $65 billion into constructing a 4 million square foot data center campus dedicated to AI development—a figure 70% higher than Wall Street analysts had anticipated. This aggressive spending strategy has seen Meta, alongside Microsoft, allocating roughly 30% of their annual revenues to capital expenditures.

Mary Long [00:32]:
"Meta plans to continue increasing their capital expenditure and invest between 60 to 65 billion dollars into building a 4 million square foot data center campus to energize AI development. That's about 70% higher than what Wall street analysts were expecting that CapEx number to be."

Asit Sharma provides a nuanced perspective on this strategy, suggesting that while such high expenditures may appear reckless, they reflect a calculated move to secure future business capacity and fend off competition from other tech titans.

Asit Sharma [01:26]:
"These people have much better view into what the future looks like than your average Joe. And they balance these competing objectives every day in committing this capital."

Sharma emphasizes that Meta and Microsoft are not only responding to current demand but are also preemptively expanding their infrastructure to capture future market opportunities and prevent competitors from eroding their market share.

The Rise of Deep Seek: A New Contender in AI

The conversation takes an intriguing turn as Mary introduces Deep Seek, a burgeoning competitor in the AI landscape. Asit Sharma compares Deep Seek's innovative approach to AI model training with the pioneering efforts of Google's DeepMind product, AlphaZero.

Asit Sharma [03:55]:
"They totally removed some of the processes of training their model that we take for granted... The results were sort of surprising. It reminds me a lot of the early days of AlphaZero."

Deep Seek distinguishes itself by enhancing model efficiency, potentially outpacing established players like Nvidia. Sharma warns that such advancements could pose significant challenges to existing hardware suppliers if Deep Seek continues to innovate.

Asit Sharma [04:30]:
"Someone could build a model that's twice as efficient as what DeepSeek is offering us and perhaps further signal down the road that somebody's going to get hurt and all this build out on their income statement in the future."

Despite acknowledging Nvidia's robust position, Sharma advises investors to remain vigilant, recognizing that the AI development landscape is still evolving with room for disruptive innovations.

AI Assistants: Meta vs. ChatGPT

Mary raises a critical point comparing Meta's integrated AI assistant with standalone services like ChatGPT, questioning whether Meta's approach offers a fair comparison.

Mary Long [05:56]:
"Zuckerberg on the earnings call predicted that this is going to be the year when a highly intelligent and personalized AI assistant reaches more than a billion people... is this a bit of an apples and oranges comparison?"

Asit Sharma concurs, elaborating on the inherent differences. Meta's AI assistant is embedded within its ecosystem—accessed through platforms like Facebook and Instagram—thereby reaching users organically as part of their daily app usage. In contrast, ChatGPT requires users to seek it out independently.

Asit Sharma [06:39]:
"Meta AI exists on different platforms within Facebook's family of apps. So it's there... whereas ChatGPT started from nothing."

This integration strategy positions Meta to potentially increase user engagement across its platforms, leveraging its existing user base to rapidly scale its AI assistant's adoption.

The Future of AI Glasses: Prediction or Possibility?

A highlight of the episode is Zuckerberg's bold prediction that 2025 will be the year of AI glasses, envisioning them as the next major computing platform. However, his statement oscillates between certainty and uncertainty, prompting questions about the feasibility and timeline of such a technological leap.

Mary Long [08:02]:
"Zuckerberg said, 'This will be the year when a highly intelligent and personalized AI assistant reaches more than a billion people... or it's going to be a longer grind.'"

Asit Sharma analyzes this prediction, noting that while there are early signs of revenue from Meta's Reality Labs (which manufactures these glasses), adoption rates are still modest. Sharma suggests that reaching a critical mass of 5 to 10 million units is essential for significant market impact, and if Meta can achieve this within the year, it could indeed be transformative.

Asit Sharma [09:56]:
"If we can have a sight line to 5 or 10 million devices on an annualized basis, in the near future, this is going to be a little bit of a game changer."

Reality Labs and Meta’s Metaverse Ambitions: A Skeptical View

Mary shifts focus to Meta's Reality Labs, which reported a revenue of $1.08 billion but suffered a loss of $5 billion in the fourth quarter and nearly $18 billion for the year. This prompts a critical assessment of Reality Labs' role in Meta's long-term growth strategy.

Mary Long [10:48]:
"Another Zuckerberg prediction we got is that 2025 will be the year of the AI glasses... how much do you expect Reality Labs to contribute to Meta's long term growth if all goes as planned?"

Asit Sharma is skeptical, dismissing the notion that this year will be pivotal for the Metaverse.

Asit Sharma [10:59]:
"How many times have we heard it'll be a pivotal year from the Metaverse... This is just about rallying the troops."

He criticizes Meta's consistent investment in Reality Labs without tangible returns, suggesting that while the technology developed may eventually benefit Meta's broader product ecosystem, Reality Labs itself is unlikely to become a materially positive segment in the near future.

Asit Sharma [13:19]:
"But are they anywhere close to having some business unit result that's going to be material and directly traceable to this? No, not anytime soon."

Microsoft's Thriving AI Business

The discussion transitions to Microsoft's impressive performance in its AI division, which has surpassed a $13 billion annual run rate, marking a 175% year-over-year growth. Unlike Meta, Microsoft's AI business is distinct from its cloud operations, encompassing various AI-driven products and services.

Mary Long [13:58]:
"Microsoft's AI business has at this point surpassed an annual run rate of $13 billion. That's up 175% year over year."

Asit Sharma breaks down Microsoft's AI offerings, which include partnerships with OpenAI, AI-integrated products like Copilot in Microsoft 365 and GitHub, and the infrastructure layer provided by Microsoft Azure. This diversified approach has positioned Microsoft as a leader in monetizing AI technologies.

Asit Sharma [14:00]:
"This is everything that Microsoft can trace to their AI proficiency and the AI services they offer... they're saying that's 13 billion bucks a year, up almost 175% year over year. Not bad."

When asked if other tech giants are generating similar AI revenues, Sharma acknowledges that while companies like ServiceNow, Amazon Web Services, and Alphabet are also in the billions, Microsoft's transparent reporting sets it apart.

Asit Sharma [15:24]:
"Maybe it's because they're out in front, but I wouldn't be surprised if we hear within the next year some more disclosures from tech giants about the actual run rate of their AI related revenue."

Azure Cloud Revenue: A Minor Miss with Long-Term Implications

Mary addresses concerns from Wall Street regarding Microsoft's Azure cloud revenue, which grew by 31% in the most recent quarter—falling slightly short of analysts' expectations of 32%.

Mary Long [16:25]:
"If 31% isn't quite hitting the mark, what should long term investors make of that?"

Asit Sharma dismisses the minor miss as inconsequential for long-term investors, emphasizing that Azure's robust performance still underscores the value of Microsoft's ongoing capital investments in cloud and AI infrastructure.

Asit Sharma [17:09]:
"I don't think it matters anything to the long term investor. The bigger story here is that... this is Microsoft's cloud infrastructure."

Decoding Microsoft's "Fungible Fleet"

In a lighter but informative segment, Mary brings up a term from Microsoft's earnings call—"fungible fleet"—used by CEO Satya Nadella in relation to Microsoft's data centers. Asit Sharma demystifies the jargon, explaining that it refers to the flexibility and adaptability of Microsoft's infrastructure to integrate various AI models and technologies.

Mary Long [17:58]:
"Nadella referred to the network of data centers Microsoft is building in his answer to this question as a fungible fleet. That phrase stuck out to me because honestly, I have no idea what it means."

Asit Sharma [18:26]:
"A fungible fleet is... the data centers and the combination of hardware and software that exists in those data centers to serve up AI... it's about flexibility... letting the best solution come."

This flexibility ensures that Microsoft can seamlessly adopt newer, more efficient AI models and hardware without being constrained by existing commitments, thereby maintaining its competitive edge.

Conclusion

The episode wraps up with Mary and Asit acknowledging the depth and complexity of big tech earnings, emphasizing the importance of understanding the strategic moves behind the numbers. They reiterate that while Meta's ambitious AI investments come with significant risks and skepticism, Microsoft's measured and diversified approach to AI presents a more promising outlook for long-term investors.

Mary Long [21:09]:
"Big tech earnings in particular bring a lot of stuff to break down, so we wanted to spend a bit more time covering today's news."

Listeners are reminded to consider the discussions thoughtfully and not make investment decisions based solely on podcast content.

Key Takeaways

• Meta's AI Spending: Meta's substantial investment in AI infrastructure signifies a strategic bet on AI's centrality to future business operations but raises questions about immediate returns, especially concerning Reality Labs.

• Emergence of Deep Seek: New players like Deep Seek introduce innovative AI models that could disrupt existing markets and challenge established hardware suppliers like Nvidia.

• Microsoft's AI Success: Microsoft's diversified AI strategy, encompassing partnerships, integrated products, and robust infrastructure, positions it strongly in the AI revenue landscape.

• Azure's Minor Growth Miss: Despite a slight shortfall in Azure's growth expectations, Microsoft's long-term investment in cloud and AI infrastructure remains promising.

• Understanding Industry Jargon: Terms like "fungible fleet" highlight the importance of comprehending industry-specific language to fully grasp a company's strategic initiatives.

Note: This summary is intended for informational purposes only and should not be construed as investment advice. Always conduct thorough research or consult with a financial advisor before making investment decisions.