As long as the ad revenue continues to flow into Meta properties such as Facebook and Instagram, CEO Mark Zuckerberg is going to invest aggressively in an AI buildout. "I think it makes sense to go for it," he said.
That's the message from Meta following its first quarter results, which landed after the company launched its powerful large language model (LLM) Llama 3. Meta reported first quarter net income of $12.37 billion, or $4.71 a share, with revenue of $36.46 billion, up 27%. Meta said its second quarter revenue will be $36.5 billion to $39 billion. The real zinger was that Meta said its 2024 capital expenditures would be $35 billion to $40 billion, above previous guidance of $30 billion to $37 billion.
"Meta needs to keep investing into AI so it keeps the consumers and clicks that drive ad revenue. If you lose the traffic because AI is better somewhere else it will be hard for Meta to keep the flywheel going," said Constellation Research analyst Holger Mueller. "And the foe is Google, which has a similar spend, but a lead when it comes to custom algorithms on custom silicon. Zuckerberg has no alternative but to spend and stay relevant. Meta must stay ahead of its ad competitors Google, and then Microsoft as well."
Needless to say, that capital expenditure guidance led to a bevy of analyst questions about Meta's AI strategy and potential. Zuckerberg was happy to elaborate. Here's what he had to say about Meta's AI buildout and the long-term view. The upshot is that Meta intends to be an AI leader.
Meta AI as helper. Zuckerberg said Meta AI will be an assistant across the company's apps and glasses that will connect people via APIs for everything from commerce to customer support as well as coding and development. The initial rollout of Meta AI has had "tens of millions of people" trying it. "We believe that Meta AI with Llama 3 is now the most intelligent AI assistant that you can freely use. And now that we have the superior quality product we are making it easier for lots of people to use them within WhatsApp, Messenger, Instagram, and Facebook," said Zuckerberg.
- Meta launches Llama 3, wide availability with more versions on deck
- Meta launches latest chip for AI workloads
Further model improvements. Meta has said it is working on a 400 billion parameter model that is still in training. "I expect our models are just going to improve further from open-source contributions," he said. "We have the talent, data and ability to scale infrastructure to build the world's leading AI models and services. And this leads me to believe that we should invest significantly more over the coming years to build even more advanced models and the largest scale AI services in the world," said Zuckerberg.
This AI investment isn't cheap. Zuckerberg said Meta will operate the rest of its business as efficiently as possible and will shift those savings to AI investments. Nevertheless, Meta is going to spend heavily. "We will still grow our investment envelope meaningfully before we make much revenue from some of these new products," he said.
Meta CFO Susan Li said:
"As we develop more advanced and compute intensive recommendation models, and scale capacity for our generative AI training and inference needs, we expect that having sufficient infrastructure capacity will be critical to realizing many of these opportunities. As a result, we expect that we will invest significantly more in infrastructure over the coming years."
Investors aren't going to like the AI investment, but Meta is used to it. The company has "historically seen a lot of volatility in our stock during this phase of our product playbook, where we're investing in scaling and new products but aren't yet monetizing it," said Zuckerberg. It has happened before with Reels, news feed and the transition to mobile. APIs are likely to be a profitable AI service for Meta. "Building a leading AI will also be a larger undertaking than the other experiences we've added to our apps, and this is likely going to take several years," he added.
The payoff is there. Zuckerberg said:
"Once our new AI services reach scale, we have a strong track record of monetizing them effectively. There are several ways to build a massive business here, including scaling business messaging, introducing ads or paid content into AI interactions. And enabling people to pay to use bigger AI models and access more compute. And on top of that. AI is already helping us improve app engagement which naturally leads to seeing more ads and improving ads directly to deliver more value."
Zuckerberg said that 30% of the posts on Facebook's feed are served by AI and 50% of Instagram content is AI recommended.
The cost efficiency of model training will be a priority. Zuckerberg said open-source contributions will likely enable efficiency gains as will Meta's proprietary accelerator chips. The goal is to run AI workloads on a less expensive stack.
There's an AI and augmented reality intersection. "Glasses are the ideal device for an AI assistant because you can let them see what you see and hear what you hear. So, they have full context on what's going on around you, as they help you with whatever you're trying to do," said Zuckerberg.
And the time is now. Zuckerberg said:
"With the latest models, we're not just building good AI models that are that are going to be capable of building some new goods, social and commerce products. I actually think we're at a place where we've shown that we can build leading models and be the leading AI company in the world, and that opens up a lot of additional opportunities."
The future. Zuckerberg said:
"I think that the next phase for a lot of these things are handling more complex tasks and becoming more like agents rather than just chatbots. You're going to give an agent something to do like an intent or a goal. Then it goes off and actually performs many queries on its own in the background in order to help accomplish your goal. Whether that goal is researching something online, or eventually finding the right thing that you're looking to buy. I think people don't even realize that they will be able to ask computers to do it for them.
I think the opportunity is really big. So, it makes sense to go for it. And we're going to, and I think it's going to be a really good long-term investment."
