Zuckerberg Says Billions Will Rely on Personal AI Agents

A day after blasting OpenAI and Anthropic for hoarding information, the drama surrounding Meta’s Q2 earnings call centered around how he would follow up. Listeners didn’t have to wait long for the answer. A day after blasting OpenAI and Anthropic for hoarding information, the drama surrounding Meta’s Q2 earnings call centered around how he would follow up. Listeners didn’t have to wait long for the answer. “It’s extremely unlikely, if you look out five years from now, that you don’t have billions of people with a personal agent that understands your goals,” Meta CEO Mark Zuckerberg said on the company’s second quarter earnings call Wednesday (July 29), describing software “working on your behalf 24/7” across health, finances, relationships and careers. That vision, for every one of Meta’s 3.6 billion daily users to have an assistant that never clocks out dominated Meta’s earnings call. Zuckerberg spent his opening remarks describing a future where personal agents handle everything from health to finances, and the rest of the call built directly on that premise: how Meta pays for it, what stands in the way, and why the company believes it will win anyway. The vision runs into a hard constraint almost immediately: supply. “There’s just nowhere near enough compute for all the demand,” Zuckerberg said, explaining that Meta is fielding offers to sell its own capacity at a premium but prefers to build intelligence on top of it instead, where the margins run higher. That scarcity is why Meta is racing to stand up the agent layer now, even in pieces. Meta Superintelligence Labs shipped its Muse Spark 1.1 and Muse Image models this quarter, and daily interactions with the Meta AI assistant have climbed sharply since it’s rebuild. Business agents are already live for more than a million businesses on WhatsApp and Messenger, handling customer inquiries and completed sales without a person in the loop. Zuckerberg pointed to Brazilian rental company Movida as the model he wants to replicate: a WhatsApp agent that runs the entire booking flow start to finish, resolving the vast majority of conversations without human help. That is the pattern Meta is betting will scale to billions of people next. Zuckerberg’s Case for Why Meta Wins the AI Bet Scarce compute raises an obvious question: Why should Meta, rather than a rival, end up owning this? Zuckerberg’s answer rests on distribution and the flywheel it creates. “When we have a product and a format that works, I would go as far as to say that I think we’re probably the best company in the world at scaling those experiences to billions of people,” he said, describing how usage data feeds back into the models that improve the products. He did not pretend the wager was small. “I get that this is a big investment and it’s a big bet,” Zuckerberg said. “My personal bet is that the people who invest in this are going to be rewarded and feel very good over time.” That confidence is the thesis the entire call built toward: an ad business strong enough to fund an all-in push into personal agents, aimed at a future Zuckerberg is convinced is coming regardless of who builds it. What Else Stood Out Meta hired Kunal Shah , founder of Indian payments company CRED, as its new head of WhatsApp. Zuckerberg said Shah built one of India’s largest payment companies. Instagram reached 2 billion daily active users. Threads crossed 500 million monthly actives, which Zuckerberg called the fastest-growing conversation app ever. WhatsApp peaked at 30 million messages sent per second during the World Cup final, an all-time record for the platform. AI-driven ranki...