Meta is accelerating its development and launch of new consumer apps by leveraging advancements in artificial intelligence, particularly large language models (LLMs). During its Q2 earnings call, CEO Mark Zuckerberg highlighted recent app releases such as a Marketplace sellers app, a Facebook Groups app called Forum, a gaming app named Vibe Coded, and an AI-driven bedtime story experiment. He pointed out that AI is enabling Meta to speed up product development and improve its recommendation systems, making it significantly easier to test and scale new app ideas.

Historically, Meta has struggled to create successful standalone social apps beyond its core platforms, with past initiatives like Creative Labs in the mid-2010s and the NPE Team in the early 2020s producing multiple experimental apps that ultimately failed and were discontinued. However, the current wave of AI technology is providing new momentum, exemplified by the rapid growth of Threads, Meta’s Twitter competitor, which has reached 500 million monthly users. Threads’ expansion benefited from AI-powered content recommendations that improve ranking and user engagement.

Meta's AI capabilities extend beyond content curation; they are also enhancing engineering workflows by automating quality assessments, trend detection, and ranking tests. CFO Susan Li explained that all Instagram Reels and Feed posts are now processed via LLMs for topic and tone analysis, refining personalized recommendations. The company is also working on recommendation systems native to LLMs, which could help new apps reach scale more efficiently.

Despite investor focus on AI spending and enterprise ambitions during the recent earnings call, Zuckerberg assured that Meta has more consumer products launching soon. The combination of AI-driven product innovation and leveraging existing user bases is positioning Meta to revive its efforts in expanding beyond Facebook and Instagram with fresh, consumer-focused applications in the near future.