• Zuckerberg says “superintelligence is in sight” as Meta pours billions into AI.
• Company moves its AI focus away from competing with ChatGPT in productivity.

Meta CEO Mark Zuckerberg has outlined what he calls “personal superintelligence,” marking a shift in the company’s AI strategy. Technology journalist Alex Heath from The Verge, who has interviewed Zuckerberg several times, believes the plan signals that Meta is stepping back from trying to directly compete with ChatGPT in productivity tools. Instead, the focus is shifting to Meta’s core motive: to keep people engaged on its platforms.

Over the past year, Zuckerberg promoted Meta’s AI assistant heavily in Facebook, Instagram, and WhatsApp in an attempt to slow ChatGPT’s momentum. But ChatGPT’s strong early lead, reputation as a productivity tool, and ability to help users complete tasks efficiently have kept it ahead. Heath says Zuckerberg appears to have realised that copying that approach isn’t working.

In his “Personal Superintelligence Manifesto,” Zuckerberg predicts that as AI boosts productivity, people will spend less time using productivity software and more time on creative and social activities. He envisions an AI that understands each user, their goals, and how to help them achieve them. While companies like OpenAI, Google, and Anthropic aim to build AI systems that take over more work, Meta wants to use AI to help fill the extra time people gain from increased productivity.

Chris Cox, Meta’s chief product officer, told employees at an all-hands meeting last month that the company will concentrate its AI efforts on entertainment, social connections, and lifestyle features rather than productivity. Heath expects this could lead to AI-powered changes to Meta’s content recommendations, ad targeting, and Reels video generation, along with interactive AI characters designed to keep users engaged longer.

The idea of “personal superintelligence” was first coined by Noam Shazeer, co-founder of Character.AI, who had considered joining Meta before returning to Google last year.

Billions for superintelligence

Zuckerberg says “superintelligence is now in sight” and claims Meta’s AI systems have started improving themselves, though progress is still slow. He did not explain what would distinguish superintelligence from standard AI but acknowledged that it will bring new safety risks. He emphasised that Meta will need to be cautious, especially with open-source releases.

Meta’s spending on AI is massive. For months, the company has been building data centres, acquiring AI startups, and recruiting top researchers, often from competitors. In the second quarter of 2025, total costs reached $27.07 billion, up 12% from the same period last year, with $17.01 billion going toward capital expenditures.

The company now expects total expenses in 2025 to be between $114 billion and $118 billion, with $66 billion to $72 billion in capital spending – higher than earlier forecasts. Infrastructure growth will account for the largest increase, followed by hiring more technical staff. Plans for 2026 point to even higher spending.

Despite the scale of investment, Wall Street has responded positively. Meta reported $7.14 in earnings per share on $47.52 billion in revenue for the second quarter, beating expectations of $5.92 and $44.8 billion, respectively. Stock jumped 10% after the results, another quarter where Meta has exceeded financial forecasts despite heavy AI spending.

Zuckerberg said Meta’s goal is to bring personal superintelligence to everyone, in contrast to rivals that focus on automating as much work as possible. He described the rest of this decade as a decisive period for shaping AI’s future – whether it will empower individuals or replace human roles.

Recruiting and infrastructure

Meta has been building its new superintelligence labs team by attracting talent from Apple, GitHub, and AI startups, often with large pay packages. One reported offer exceeded $200 million. The company also invested $14.3 billion in Scale AI for a 49% stake, bringing its CEO, Alexandr Wang, on board as chief AI officer.

Mike Proulx, research director at Forrester, says Meta’s willingness to spend heavily is giving it an edge in recruiting top AI experts and building the infrastructure to support its ambitions. “Money talks and Meta has plenty of it,” he said.

Reality Labs, Meta’s division for hardware like AI glasses, remains a small part of the business, with $370 million in revenue last quarter. But Zuckerberg compared AI glasses to contact lenses, saying people without them could be at a “cognitive disadvantage.”

Advertising remains Meta’s largest revenue source, bringing in $46.6 billion in the second quarter, up from $38.3 billion a year earlier. CFO Susan Li said WhatsApp ads are unlikely to contribute meaningfully to ad revenue growth for several years, as the platform targets lower-monetising markets and offers less data for ad targeting than Facebook or Instagram.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.