© metamorworks / Shutterstock.com
Shares of Meta Platforms (NASDAQ:META | META Price Prediction) have been clobbered since the firm pulled the curtain on its underwhelming quarterly earnings results. The numbers themselves were okay (28% revenue growth is nothing to sneeze at), but compared to the other hyperscaler firms, the results were actually quite uninspiring, with a lot to be desired. Add heavy spend into the equation, and it’s not a mystery as to why investors rotated out of the name so violently last week.
One bad quarter doesn’t signify that all is lost with Meta, though, especially as the company continues to spend aggressively to catch up in the hyperscaler race while also funneling considerable sums into its superintelligence team, which now actually has something quite impressive to show for its many years of hard work with Muse Spark, a model that might actually convince some users to pay for.
As the company shifts gears a bit from using its own AI compute to selling compute to others via Meta Compute while looking to play in the same arena as Anthropic, OpenAI, and Alphabet’s (NASDAQ:GOOG) Google with its new closed-source large language model Muse Spark, questions linger as to how the company’s AI strategy stands to evolve from here.
Yes, Meta is spending a lot. But it’s probably going to be money well spent
While Meta’s seemingly worrisome spend to fund its AI plans was a primary reason to hit the sell button after the second-quarter reveal, I do think that to sell Meta here is to forego a dirt-cheap stock that’s well on track to win in the AI race as it looks to aggressively move forward in a way that could disrupt its fellow hyperscaler peers.
At the end of the day, Meta might be less far along in its CapEx cycle, but I do think it will make up for this with heavy spending and creative decisions made internally. Whether that’s an AI Zuckerberg or making use of Meta’s trove of internal data, I do think the firm is more than willing to add an exclamation point when it comes to its AI plan.
The major positive for Meta, at least in my opinion, is that Meta is entering the hyperscaler game at the best possible moment. It’s got the AI-native, AI-first mindset before the shovel goes into the ground — an advantage that I think much of Wall Street is heavily discounting, especially following a discouraging quarterly showing.
Without the capital drag of legacy cloud infrastructure and some ridiculously aggressive moves to move at full speed in the AI race, my guess is that it’s just a matter of time before Meta Compute becomes the fastest, shiniest new racecar that’s driving in the AI race.
A powerful frontier AI innovator and new hyperscaler that’s going for a bargain
As Meta looks to take control of the stack with its hardware initiatives while exploring the possibilities within the consumer and business realms with its profoundly powerful and rapidly advancing new model, I certainly wouldn’t want to bet against the firm while it’s down and out, trading for 18.6 times forward price-to-earnings (P/E) — a multiple that I believe vastly underestimates the company’s AI vision as well as the sheer aggression of Mark Zuckerberg.
With a powerful digital ads cash engine behind the firm and the potential for high-margin cloud growth to take things into overdrive, I certainly wouldn’t overlook Meta’s shares, even though there are more exciting things in the hyperscaler world to look to right now. Add the wild card of Muse Spark into the equation and the potential for deeper cuts to Reality Labs, an effort I’m sure Wall Street would applaud, and I see Meta as having very easy levers to pull to get its stock marching higher again.
Dismiss Muse Spark, if you will, but version 1.1 is a pretty big deal. It’s a powerful model with seriously impressive agentic capabilities, with benchmarks that suggest it’s not all too far from the very front of the pack. Meta made a statement with this release, and it’s going to be interesting to see how much better the next iteration will be as Meta’s Superintelligence Labs looks to win at any cost.
Contact [email protected] for any questions or corrections.
