Meta CEO Mark Zuckerberg
Brendan Smilalowski/AFP via Getty, Tyler Le/BI

  • Mark Zuckerberg has recently become “more ambitious” about Meta’s ability to win in AI.
  • He now plans to spend around $40 billion this year, largely on AI investments.
  • In the next few years, Zuckerberg sees three ways AI can become a “massive business” for Meta.

Mark Zuckerberg is now convinced that Meta is a top AI company, and he even knows how the technology will become a significant source of profit in the years ahead.

With the recent release of Llama 3, Meta’s latest AI model, Zuckerberg said he became “more ambitious and optimistic on AI” and his company’s ability to deliver on the tech.

He made it clear during a Wednesday earnings call with analysts that he intends to “invest significantly more over the coming years to build even more advanced models and the largest scale AI services in the world.”

“With the latest models, we’re not just building good AI models that are capable of building some new good social and commerce products,” the CEO told analysts. “I actually think we’re in 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 beyond just the ones that are the most obvious for us.”

Heavy spending again

Such ambition does not come cheap. Meta increased its guidance on capital expenditure for this year, saying it now plans to spend between $35 billion and $40 billion, largely on AI investments. Its stock slumped 16% in after-hours trading.

The last time Zuckerberg got excited about a new technology (the metaverse), Meta spent wildly and freaked investors out. The stock collapsed and didn’t recover until the company embarked on a “Year of Efficiency” marked by mass layoffs and a more business-minded CEO.

Zuckerberg on Wednesday made a concerted effort to head off Wall Street panic that his new AI enthusiasm is lacking in business acumen.

He sees “several ways” generative AI can make money, and laid out three specific paths to this becoming “a massive business” for Meta. Although getting there is a “long-term” prospect, he warned.

“Business messaging”

One of the ways AI can make money is by building up “business messaging,” where companies pay Meta for generative AI tools, such as services that support automated interactions with users and customers. Zuckerberg envisions Meta’s AI moving beyond just being a chatbot and becoming an AI “agent” that handles more complex tasks, and processes multiple queries to solve user problems, instead of coming back instantly with rote answers.

Revenue from AI business messaging is “one of the nearer term opportunities,” Zuckerberg said. While it may not become a reality this year, he noted that it’s less than five years away. The immediate goal on this front is to “get many hundreds of millions or billions of people to use Meta AI as a core part of what they do,” he explained.

Ads appearing in “AI interactions”

Another way generative AI could make money for Meta is by “introducing ads or paid content into AI interactions,” Zuckerberg said. Although brands and companies paying for products to show up in generative AI results is not yet the standard for AI chatbots, Meta’s entire business is effectively driven by selling digital advertising. Inserting ads into its social and messaging products is at the core of Meta as a company.

AI is already being more widely deployed by Meta in its newer “unconnected content” algorithm for social content recommendations, which Zuckerberg said is leading to more app engagement. That, in turn, leads to more people seeing more ads. Right now, 30% of the content Facebook users see is recommended by AI, and the same goes for 50% of the content seen by Instagram users, he said.

Selling access to AI models

A third distinct way Meta may make money from AI is by selling access to models as they get larger. “Enabling people to pay to use bigger AI models and access more compute,” as Zuckerberg put it on Wednesday.

Right now, Llama 3 and the Meta’s other large language models are freely available to users and companies below a certain size threshold. Charging for access might be a move away from Meta’s “open source” approach here.

“So if the technology and products evolve in the way that we hope, each of those will unlock massive amounts of value for people and business for us over time,” Zuckerberg said. “I think it makes sense to go for it, and we’re going to.”



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *