Meta Platforms Inc. is sending mixed signals regarding its massive artificial intelligence strategy. While the tech giant has quietly entered the consumer generative gaming market with a brand new application, its Chief Executive Officer, Mark Zuckerberg, has internally admitted that the company’s broader, multi billion dollar AI restructuring and technology timeline are falling short of expectations.
On the consumer side, Meta soft launched a mobile application called Pocket on both the Apple App Store and Google Play Store. The app was released quietly without an official company announcement, but it was quickly spotted by reverse engineers and app intelligence firms. Pocket allows everyday users to build and share fully functional, interactive mini games and mini applications using nothing but plain text descriptions a concept the tech community refers to as “vibe-coding.”
Instead of writing complex code, users type out what they want to experience, and the AI generates what Meta officially calls a “gizmo.” These gizmos are interactive, touch first experiences that can leverage smartphone hardware, such as touch controls, cameras, and tilt sensors. Structurally, Pocket functions like a social network, featuring a vertical discovery feed where users can play games built by others or “remix” them to add their own creative twist. The app is built directly on Meta’s strategic acquisition of the startup team behind Gizmo, an independent AI-gaming platform.
However, behind the scenes at Meta’s headquarters, the picture is far more complex. During an internal town hall meeting, Mark Zuckerberg candidly addressed employees, admitting that the trajectory of Meta’s highly anticipated autonomous AI agents has been underwhelming.
According to recordings of the meeting, Zuckerberg stated bluntly: “In retrospect, the trajectory of the agentic development over at least the last four months hasn’t really accelerated in the way that we expected.” He further acknowledged that the internal shake ups designed to catalyze this technology have stalled, adding that the company’s bets on its new organizational structure “haven’t come to fruition yet.”
The corporate hurdles follow an aggressive internal overhaul. Meta initiated a sweeping restructuring that included cutting roughly 10% of its global workforce and reassigning approximately 7,000 employees into dedicated AI focused teams. The aggressive transition dented internal staff morale and faced pushback. Zuckerberg admitted to employees that the execution of these layoffs and team transfers was “not as clean” as leadership had originally planned, noting that senior executives ultimately misjudged the timing of the transition.
Internal planning sessions showed that Meta executives were “super optimistic” about the rapid emergence of next generation AI coding tools, such as competitor Anthropic’s Claude Code, and feared the company was not moving fast enough to adapt.
Despite the technical bottlenecks and organizational friction, Meta is projected to spend up to $145 billion on AI infrastructure. Zuckerberg remains defensive of the capital output, telling staff that he expects the social media behemoth to begin seeing more significant benefits and returns from its AI investments within the next three to six months.
Ultimately, the duality of Meta’s current position highlights a distinct industry reality. While lightweight consumer applications like Pocket prove how rapidly AI can change user generated content today, the internal friction at Meta underscores how arduous and unpredictable it remains to construct the foundational, automated AI future that Big Tech has promised investors.
