Jakarta, INTI – Meta, the parent company of Facebook and Instagram, has officially launched the standalone Meta AI app, expanding the presence of its virtual assistant beyond internal platforms like WhatsApp, Messenger, and Instagram. Powered by Meta’s latest large language model, Llama 4, the app is designed to deliver a more personalized and contextual interactive experience, positioning itself as a direct competitor to ChatGPT, Google Gemini, and other AI platforms.
Key Features: Natural Voice Interaction to AI-Driven Inspiration Feed
Meta AI introduces full-duplex technology that enables more natural, voice-based interactions. Beyond simply reading text, this AI can now respond in real-time, mimicking human-like two-way dialogue. Although currently available only in the U.S., Canada, Australia, and New Zealand, this feature marks a major step for Meta in redefining digital conversational interfaces.
In addition, Meta AI introduces the Discover Feed, allowing users to explore AI prompts and ideas from other users while sharing their own creativity. Thus, the app serves not only as a personal assistant but also as a collaborative social space for AI exploration.
Beyond its standalone version, Meta AI is also integrated into wearable devices like Ray-Ban Meta smart glasses and the meta.ai website. The app is designed to learn user habits through data connections with Facebook and Instagram accounts, enabling more relevant and contextual responses.
Massive Business Potential: Trillions of Rupiah from Generative AI Products
Behind the launch of Meta AI lies a major business ambition. According to TechCrunch, Meta is projected to generate between IDR 33 trillion and IDR 49.5 trillion in 2025 from its generative AI product line. Over the long term, this figure is expected to skyrocket to between IDR 7.5 quadrillion and IDR 23 quadrillion by 2035.
These projections come from multiple revenue streams, including Llama model licensing partnerships, customized model API launches, and potential monetization of Meta AI through ads and subscription options. Meta CEO Mark Zuckerberg has also hinted that paid premium features may be introduced in the future.
However, this bright outlook is shadowed by legal challenges. Meta is currently facing lawsuits from authors alleging that the company used their work without permission to train its AI models. While Meta’s spokesperson denies the allegations, citing “fair use” principles, court documents reveal that the company had considered purchasing legal training data worth IDR 3.3 trillion before ultimately opting to download massive amounts of pirated ebooks.
On the other hand, developing this AI requires substantial investment. Meta’s generative AI division had a budget exceeding IDR 14.8 trillion in 2024 and is expected to grow beyond IDR 16.5 trillion this year—not including the cost of infrastructure like large data centers, which are projected to be funded through capital expenditures of IDR 990 trillion to IDR 1.4 quadrillion in 2025.
Conclusion: Meta AI, an Ambitious Strategy in the Global AI Race
With the launch of the standalone Meta AI app, Meta is entering a new chapter in the race for AI dominance. The focus on voice interaction, social integration, and personalized experiences signals a strategic direction to capture the consumer AI market.
However, amid this aggressive expansion, Meta faces legal and ethical scrutiny over the data used to train its AI models. Nonetheless, with staggering revenue forecasts and massive investment commitments, Meta AI represents more than just a product launch—it is the foundation of Meta’s future in the era of artificial intelligence.
Read More : Wikipedia is ready to use AI, but still relies on editors and volunteers as the main pillars