AI assistants are about to become far more central to daily life, if Meta’s chief executive is right. Speaking at the company’s Q2 earnings call on July 29, Mark Zuckerberg predicted that billions of people will have an AI assistant working on their behalf within five years, handling everything from financial decisions to household management.
What Did Zuckerberg Actually Say About AI Assistants?
Zuckerberg described a future where the AI assistant “understands your goals and works 24/7 to achieve them, regardless of the field,” according to VnExpress KHCN’s report on the earnings call. He named four specific use cases: financial problem-solving, health decisions, personal relationships, and household tasks.
On distribution, he singled out messaging platforms as the primary interface. “As we move toward a future where everyone interacts with multiple AI assistants, I think WhatsApp and other messaging platforms will become increasingly important,” he said. The company confirmed that WhatsApp is currently the leading platform through which users interact with Meta AI.
What Do the Numbers Behind Meta’s AI Bet Reveal?
The financial picture is striking. Meta reported free cash flow of $784 million for the quarter, a 91% decline from the $8.55 billion recorded in the same period the prior year, according to the company’s earnings disclosure. The company attributes the drop to heavy investment in AI infrastructure.
The day before the call, on July 28, Meta announced a partnership with BlackRock to build a $14 billion data center in El Paso, Texas. Zuckerberg told investors he believes “the margin from selling AI will be significantly higher than selling computing power directly,” and described the personal AI assistant as “the foundation for the next wave of products and revenue” at Meta.
Is the AI Assistant Market Convinced by Meta?
Not entirely. Despite the scale of investment, Business Insider reported (as cited by VnExpress KHCN) that Meta has not yet persuaded customers or developers that its models are top-tier, comparable to offerings from OpenAI, Google, or Anthropic.
Chamath Palihapitiya, a former Facebook executive, told Axios that Meta is “unlikely” to catch up with rivals in the generative AI (AI that produces text, images, or code on demand) race. “I don’t understand the underlying dynamics to explain why they’ve failed so miserably,” he said, adding: “But currently, the ambition has not materialized.”
What Has Been Happening Inside Meta?
The AI push has coincided with significant internal turbulence. In May, Meta laid off roughly 10% of its workforce, approximately 8,000 employees, and restructured around 7,000 other positions, according to VnExpress KHCN. That same month, employees distributed flyers at multiple U.S. offices expressing frustration over the layoffs and a computer monitoring program introduced by management.
In late June, Meta paused the monitoring program to investigate a suspected data leak. AFP reported in early July that many employees remained in a state of “anxiety mixed with anger.”
What This Means for AI-Search Visibility
Here is the simplest way to frame the business risk: if people start asking an AI assistant “who should I use for X?” instead of typing it into Google, brands that are not in the AI’s answer simply do not exist for that question.
Picture a concrete example. Someone asks their phone’s AI assistant to recommend a local physiotherapist. The assistant does not show a list of links. It names one or two options it has learned to trust. Businesses not represented in the data those assistants draw on, through review platforms, professional directories, or authoritative web content, receive no referral from that exchange at all.
The WhatsApp signal from this earnings call deserves specific attention. Meta confirmed WhatsApp as the top interface for Meta AI interaction today. The standard SEO logic of “build a visible website and distribution follows” does not transfer cleanly to an AI assistant embedded in a private messaging app.
The sources here cover Zuckerberg’s vision and Meta’s financials. One dimension the reporting does not address: distribution scale may prove more decisive than model capability in the near term. Even if Meta AI lags OpenAI in benchmark rankings today, an AI assistant operating at WhatsApp’s volume still shapes a meaningful share of purchasing and service decisions. A less capable AI answering far more questions influences more outcomes than a superior AI that fewer people reach.
Hingewise’s assessment: the most actionable signal in this earnings call is not the five-year forecast but the confirmation that WhatsApp is already the primary Meta AI touchpoint today. For brands that depend on social and messaging traffic, the question of how an AI assistant finds, surfaces, and summarizes their information has moved from theoretical to operational.
What to Check Before AI Assistants Scale Further
- Identify which platforms your customers already use to ask questions: messaging apps, voice assistants, AI chatbots.
- Test what a major AI assistant currently says when asked about your business category or the core problem you solve.
- Check whether competitors appear in those AI responses and you do not, then investigate what content or listings they have that you lack.
- Audit whether your core product and service information is published in a format AI systems can read easily: clear headings, direct answers to common questions, structured data markup where applicable.
- Verify your business is listed accurately in directories and platforms that AI assistants commonly draw from: review sites, health directories, professional registries.
- Assess whether your content answers specific questions directly, not just describes what you offer in general terms.
The gap between Zuckerberg’s five-year prediction and current market reality remains wide. Whether Meta narrows the capability gap with OpenAI and Anthropic, or whether WhatsApp’s distribution advantage proves more decisive than model rankings, will be a defining question for the AI assistant market over the coming years.
Sources: VnExpress KHCN, citing Meta Q2 earnings call (July 29), Business Insider, Axios, and AFP. Original report (Vietnamese).
Lam Nguyen · Hingewise