Meta CEO Mark Zuckerberg Just Assembled a “Super Intelligence Avengers” Team That Could Totally Change the Game in Artificial Intelligence (AI). Here’s Why That Makes Meta a “Must-Own” AI Stock.

by

Many investors are closely watching the developments in the field of artificial intelligence (AI), as the leading seven firms are investing significant resources in a competitive race to lead in AI technology, which could yield substantial financial rewards.

Over the past few weeks, the CEO of Meta Platforms (META), Mark Zuckerberg, has taken monumental steps, investing vast sums of money in both human talent and computing infrastructure. These investments surpass the extravagant spending standards even of today’s leading AI companies.

Some people might understand the potential consequences of these actions, but others in the broader investment world might not fully appreciate their significance yet.

Loading widget...

Zuck throws down the gauntlet

Over the past month or so, Zuckerberg has:

  1. Purchased 49% of data-labeling leader Scale AI at a $28 billion valuation, bringing in Scale’s CEO Alexandr Wang and top leadership.
  2. Hired top AI talent in addition to Wang to create a “Super-Intelligence Team” from several leading AI and tech rivals, totaling about 50 researchers, by offering multiples more than other companies, with some offers rumored to be as much as $200 million or more.
  3. Notable poached talent includes Nat Friedman, the former GitHub CEO; Daniel Gross, who was CEO and co-founder of SSI, Ilya Sustkever’s current start-up (Sustkever was a co-founder of OpenAI); Ruoming Pang, the head of Apple‘s AI division; as well as Lucas Beyer, Alexander Kolesnikov, and Xiaohua Zhai from OpenAI.

On infrastructure investments, Zuckerberg also shed light on massive upcoming projects:

  1. In a Threads post, Zuckerberg said Meta was going to invest “hundreds of billions of dollars” in AI superclusters.
  2. This includes the industry’s first 1GW supercluster, which Meta is calling Prometheus and should come online in 2026.
  3. Zuckerberg also said this will be just the first of multiple GW-plus superclusters, including Hyperion, which will eventually scale up to 5 GW over several years, and encompass a data center almost the size of Manhattan.

How does all this spending pay off?

As a fervent observer, I’ve often pondered over the sudden surge in spending by Zuckerberg. Was it a strategic offense or defense he was playing? Well, my guess would be that it’s probably a mix of both!

Mark Zuckerberg recently expressed that Meta’s goal is to develop “high-level intelligence,” a term similar to what was previously known as Artificial General Intelligence (AGI). The idea of high-level intelligence has sparked much discussion about whether AI can actually attain this level, but it seems that Zuckerberg thinks it’s possible and may be achieved within the coming years.

In a recent interview with tech magazine The Information, Zuckerberg said:

There’s a heated discussion within the industry currently about when super intelligence might become a reality. Frankly, it’s hard to predict whether it will happen in three, five, or seven years. However, my personal view is that we should prepare and make decisions as if it could be achievable within the next two to three years.

Zuckerberg considers that the term “super intelligence” could hold distinct meanings for Meta compared to more corporate-focused companies like The Mag Seven. For instance, Microsoft might employ AI to automate various enterprise tasks, boosting productivity. In contrast, it appears that Zuckerberg envisions Meta providing consumers with a form of “super intelligence” that enhances their daily experiences, the media they engage with, and their social interactions.

In the intriguing conversation I observed, Zuckerberg pointed out a thought-provoking perspective. He suggested that the substantial remuneration is justified given the compact size of the elite squad, ranging from 50 to 70 individuals.

In simpler terms, it seems more effective when working on this project that a smaller number of highly skilled individuals are involved. This is because it’s easier to manage and understand complex concepts with fewer minds. Therefore, there’s a significant value placed on having the best and brightest team members.

It’s clear that designing AI systems is intricate work, and even a single misstep in the system architecture can impact the overall model’s performance significantly. As reported by the AI chip blog Semianalysis, Meta’s latest large language model Llama 4 has underperformed due to several factors, including subpar data labeling – an issue that could potentially be addressed through Meta’s acquisition of Scale AI – and some architectural flaws.

It’s not surprising that Zuckerberg chooses to invest in fewer but top-tier engineers for Meta’s AI projects, given the significant impact key decisions can have on success or failure. It seems he prefers quality over quantity in this instance, which aligns with his desire to create a winning model for Meta’s AI ventures.

One issue that arises is the significant financial advantage Meta seems to hold over its competitors, particularly OpenAI, currently operating at a substantial loss. Last year alone, Meta’s primary social media advertising business generated an impressive $87.1 billion in operational earnings, albeit with a $17.7 billion deficit in its Reality Labs division. It is likely that this figure will approach the $100 billion mark this year.

Consequently, I find that Meta can match or surpass remuneration offered by competitors, and by doling out such immense wages, it’s inadvertently increasing the overall labor costs – even for companies like OpenAI. Zuckerberg further stated:

One advantage of reinforcement learning is it offers an opportunity to transform substantial resources into progressively improved services, perhaps surpassing what less resourceful or less daring competitors can achieve. I consider this a significant competitive edge. If we can master this technique effectively, and that’s the main reason we are fully committed. We’re constructing numerous data centers with multi-gigawatt capacity, which we aim to fund entirely from the company’s cash flow.

But the move may also be defensive, and isn’t without risks

Although Mark Zuckerberg’s high-stakes investment in Meta’s AI development is captivating, investors ought to consider some potential concerns. To begin with, it seems that Meta’s ambitious goal of creating an advanced AI super team may require a significant amount of groundwork, perhaps because their previous Llama 4 project didn’t meet expectations or fell behind the progress of its competitors. This suggests that Meta’s current AI leadership initiative might not be as promising as anticipated, leading some to question whether it’s wise to invest heavily in this area right now.

I’ve observed that, despite his efforts, Mark Zuckerberg wasn’t able to secure all the companies and talents he aimed for. For instance, he reportedly sought to acquire Mira Murati’s Thinking Machines and Ilya Sustkever’s SSI, but was turned down in both instances. Similarly, he extended billion-dollar offers to some of OpenAI’s top leadership, only to be rejected once more. Consequently, while Meta now boasts a highly capable AI team, it may not be the complete dream team Zuckerberg envisioned.

To summarize, it’s no secret that Meta has been known for investing heavily in distant projects without immediate returns. For instance, the Reality Labs division represents Mark Zuckerberg’s venture to establish the future computing platform through virtual reality goggles or glasses. In fact, the company rebranded itself from Facebook to Meta Platforms in 2021, signaling its dedication to this endeavor. However, by 2024, three years after the rebranding, the Reality Labs segment reported a loss of $17.7 billion, an increase from the $16.1 billion loss it experienced in 2023.

In summary, Zuckerberg didn’t explicitly define what he referred to as a “super intelligence” for everyday consumers. Although his Reality Labs division and the idea of consumer-level superintelligence might become reality someday, it’s not guaranteed – not even with Zuckerberg gathering an elite AI team. Consequently, while this month’s spending has been thrilling, investors may grow restless if Meta continues to increase its expenditures without a substantial rise in income.

And yet, the spending makes Meta a must-own stock

If a leading tech company of the present-day tech giants (referred to as The Magnificent Seven) were to achieve “super intelligence” before its competitors, it could significantly shift the power dynamics among them. Therefore, it is advisable for both young individuals and growth investors to consider investing in Meta Platforms Inc. and its peers, despite their substantial AI investments currently.

If any of these companies successfully solve a significant challenge ahead of their competitors, it’s plausible that the current top seven companies could reduce to three, two, or even just one. With his recent strategic investments, Zuckerberg is putting considerable effort into ensuring Meta is among the frontrunners aiming to be “the one” that emerges victorious in this competition.

Listen carefully for updates as Meta is due to announce its earnings by the end of this month, specifically on July 30th.

Read More

2025-07-19 14:43