Mark Zuckerberg Is Borrowing Facebook's Old Playbook to Win the AI Race

Mark Zuckerberg, the famed CEO of Meta Platforms (NASDAQ:META) dropped a 6,500-word manifesto on the company’s website Monday outlining his -- and his company’s -- expansive plans to bring free versions of artificial intelligence to billions of people around the world.

Zuckerberg goes into detail in his vision for personal AI assistants that can help people with finances, careers, hobbies, and relationships, and he predicts that AI will help create jobs rather than eliminate careers.

“We propose a philosophy based on individual empowerment as the source of prosperity, invention as the primary purpose of superintelligence, and balance of power as the foundation of safety,” he wrote. “Superintelligence will be among the most important technologies in history, so its opportunities and challenges will likely be greater than any we’ve seen in our lifetimes. We should take this very seriously.”

It’s a bold declaration. And in reading it, I think it’s also clear that Zuckerberg is returning to his old playbook from Facebook’s early days as he looks to spread Meta’s version of “personal superintelligence” worldwide.

The Facebook model

Zuckerberg came up with Facebook while he was a teenager attending Harvard University. The site was designed to help users match names to the faces of fellow students and quickly caught on, spreading to other universities and the general public.

One interesting thing at the time was that Zuckerberg focused nearly exclusively on scaling the company and driving user engagement rather than on profits. It didn’t launch its Facebook Ads platform until 2007 -- three years after the site launched. And when the company was building its mobile app in 2010, Zuckerberg was more focused on attracting 1 billion active users rather than finding a way to sell ads.

“There’s no point right now in having a massive profit,” he said. “(Ads) factor in, like, not at all.”

His focus worked. The mobile app made Facebook even more popular, and by 2013, the company was placing greater emphasis on revenue by serving targeted ads in users’ feeds.

Today, Zuckerberg’s company, which changed its name from Facebook to Meta in 2021, operates Facebook, Instagram, WhatsApp, Messenger, and Threads, and has an average of 3.5 billion daily users. Advertising accounts for nearly all of Meta’s revenue, making up 97% of the company’s $60.8 billion in the second quarter.

In short, Zuckerberg focused on making sure as many people as possible were using Facebook before he got serious about fully monetizing the product. And he’s doing the same thing with superintelligence and AI.

Bringing superintelligence to the masses

In his manifesto, Zuckerberg writes about AI and superintelligence almost as if it were a human right.

“Some argue that superintelligence itself or a small set of experts who control it should decide what is best for humanity. We disagree. The history of democracy and economics has shown that there is no single objective answer to how people define the best life, and therefore the best approach is letting people decide what matters in their own lives. We believe that delivering superintelligence to everyone is the way to answer this question.”

Zuckerberg said Meta will offer free versions of AI personal agents, charting a very different path from AI start-ups OpenAI and Anthropic, which rely on subscriptions, API usage, and enterprise contracts to raise revenue and build out their large language models. Those who want more compute will pay through what he calls a “dynamic auction mechanism” to keep prices low.

But the difference between Meta, OpenAI, and Anthropic is that Meta already has an established, powerful advertising revenue stream through its social media platforms to help it develop and fund superintelligence. That’s important because building out AI infrastructure is tremendously expensive, and that cost is already weighing down Meta Platforms’ stock.

Meta will need to keep the advertising revenue coming in as it works on superintelligence and free AI agents. But if it can scale that business as it did with Facebook, then Meta will be in a position to find ways to profit from its efforts.

And there will be plenty of chances to do so, starting with revenue for additional compute, and products such as augmented reality glasses that Zuckerberg says will be needed “to keep you present in the moment with the people you care about.” Meta could also find ways to monetize AI-generated transactions and business contacts through its apps, or use the data it collects to make its advertising efforts more effective and valuable.

For all that to happen, though, Meta needs to make personal AI agents that people want to use, which means focusing on the user experience, making them fun, and ensuring they are extremely accessible.

That’s the model that worked for Facebook two decades ago. And that’s the model Zuckerberg needs to work for Meta now.

Patrick Sanders has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Meta Platforms. The Motley Fool has a disclosure policy.