Meta Platforms (NASDAQ: META) is probably one of the most misunderstood companies out there. It is consistently judged by what it does with 5% of its business rather than how the other 95% is performing. This misunderstanding was on full display after Meta reported fantastic Q3 earnings, only for the stock to drop right after the report.
Because Meta Platforms is heavily investing in the artificial intelligence (AI) space, some investors are panicking. However, I think that’s the wrong way to look at it. Instead, I’m assessing the company to see if it’s an AI leader, as a leadership position could be worth billions of dollars.
Start Your Mornings Smarter! Wake up with Breakfast news in your inbox every market day. Sign Up For Free »
First, let’s talk about what Meta does better than anyone else. Its core business — social media sites like Facebook and Instagram — is a cash cow. It thrives off of advertising revenue, which was up 19% year over year in Q3. Of the $40.6 billion in revenue that Meta generated, $39.9 billion came from advertising on its social media sites.
Clearly, this is the main part of the business investors should be focusing on, yet it’s easy to get distracted by Meta’s side hustles.
Investors don’t have the best experiences with Meta’s side hustles. Meta spent billions of dollars building out its vision for the metaverse, only for nothing to come to fruition in that space. Now, it’s pouring a ton of money into its AI models. But Meta’s generative AI model, Llama, has actually been quite successful.
Llama has seen adoption as the building block of internal AI models for AT&T, Goldman Sachs, and Shopify. From January 2024 to July, Llama usage grew tenfold, which demonstrates rapid adoption.
Meta is looking to capitalize on this usage by developing the next generation of the model. Currently, Llama is in its 3.2 version state, but Meta is developing Llama 4. The computing power necessary to train Llama 4 versus Llama 3 is expected to be about 10 times as much.
In its quest to produce the best generative AI model, Meta conveyed to investors that they should “expect significant capital expenditures growth in 2025.” This concerned many investors as they want Meta to focus on being a cash cow and returning profits to shareholders through dividends and share repurchases.
However, that’s shortsighted thinking, and if Meta can develop a leading AI model, it will far outweigh the benefit of a dividend or share repurchase program. The stakes are too big not to have a top-notch generative AI model, and with Meta’s near-term success in having clients adopt it, investors shouldn’t think of this as the next metaverse. Instead, this could be the next Facebook.
