META Stock Analysis: How Meta Platforms Won the AI Wars

Stocks to buy

While its Cloud rivals were falling after earnings, Meta Platforms (NASDAQ:META) stock stood firm.

Microsoft (NASDAQ:MSFT) was down 5.4% on the week. Alphabet (NASDAQ:GOOG, NASDAQ:GOOGL) was down 1.2%. Amazon.com (NASDAQ:AMZN) tumbled 8.5%.

Meta Platforms, on the other hand, was up 3.9%. This includes a drop of nearly 2% on August 2, the worst day for the market in years. META reported net income of $13.46 billion, and earnings per share were $5.16, 73% ahead of a year ago. Revenue was $39 billion, up 22%.

How did they do it? More important, why were they able to do it? What does this portend for the future? Beyond the numbers, what were the secrets of META’s win in the Great Artificial Intelligence Wars of 2024?

A Global Footprint

Much of it has to do with decisions co-founder and Chief Executive Officer Mark Zuckerberg took over a decade ago.

Zuckerberg committed to building a global cloud network, before he had the cash flow to sustain the effort. Facebook created the Open Compute Project to minimize costs. It linked its data centers to customers via a network of undersea cables.

This means Meta Platform’s 25 data centers don’t have to be located near customers, in politically dangerous waters. The vast majority are in the U.S., with three in Europe and one in Singapore.

So, while rivals must rent their clouds and fiber infrastructures, META owns its infrastructure, which is built on cash flow, not debt. The balance sheet still shows just $18.4 billion in debt, against $32 billion in cash.

Firstest With the Mostest

This is as true in business as it is in battle. Business battles are won long before they’re fought. Getting resources in place well in advance is key.

Changing the company’s name from Facebook to Meta Platforms wasn’t a casual thing. Zuckerberg was deploying programming resources well ahead of the battle, before he even knew where the battle would be fought. (He thought it would be about virtual reality. It wasn’t.)

This cost shareholders dearly in 2022. Meta Platforms lost two-thirds of its value during the “tech wreck” that year, falling from $336 to a low under $100. But pre-positioning assets was the key for what followed, a five-fold increase in the market cap, now nearly $1.2 trillion.

The company is building out a network with 350,000 Nvidia (NASDAQ:NVDA) H-100 Hopper chips, in clusters of nearly 25,000 each. That’s going to be hard to match.

Get Help

The final ingredient was getting help.

META’s Large Language Model (LLM), called Llama 3.1, is open source. You can download it, call out bug fixes and make improvements under the Llama Community License.

The license is limited. Critics say it isn’t even open source. There’s also a commercial license.

This means programmers around the world, including Asia, Africa and South America, and people who don’t have money but may have talent, can learn, use and build with META software. They can’t do it with the products of Meta Platform’s rivals. That would be dangerous, its rivals say.

The Bottom Line

The Cloud changed the technology game in the 2010s.

But the Cloud isn’t what matters anymore. With enough Nvidia racks, anyone can replicate what the 5 biggest infrastructure players spent a decade building.  

The AI era will be more like the PC era that preceded it. Operating systems were the key back then. They’re how Microsoft turned IBM (NYSE:IBM) into a second-rate outfit, although few saw it at the time.

Meta Platform’s advantage in mass deployment of cloud models that anyone can use doesn’t look very large right now. However, if you’re an investor, it’s what you want to be focused on, because the AI era has just begun.

As of this writing, Dana Blankenhorn had a LONG position in MSFT and AMZN. The opinions expressed in this article are those of the writer, subject to the InvestorPlace.com Publishing Guidelines.

On the date of publication, the responsible editor did not have (either directly or indirectly) any positions in the securities mentioned in this article.