I just reviewed something quite interesting happening at Meta. It seems the company is preparing to launch entirely new AI models under the leadership of Alexandr Wang, who heads their superintelligence lab. The curious thing is that Meta is adopting a hybrid strategy: planning to release open-source versions of these models but keeping the most advanced capabilities as proprietary technology.

What catches my attention is the strategic approach behind this. While OpenAI and Anthropic focus on enterprise and government markets, Alexandr Wang is betting heavily on consumer mass adoption. His vision is to turn AI into a personal superintelligence that helps ordinary users expand their capabilities, not just serve large corporations.

The new model will likely include multimodal capabilities (text, image, video) based on previous exploration of projects like Avocado and Mango. Although Meta already has its Llama series in open source, this new development is different. Wang acknowledges that initial performance might not match top products from OpenAI or Google Gemini, but he trusts that usability, privacy, and free or low-cost access advantages will create differentiation.

From Alexandr Wang’s perspective, the era of superintelligence is near, and Meta has the resources, talent, and ambition to drive massive scientific advancement. The hybrid open-source strategy aims to attract developers without exposing itself to full security risks. It’s pragmatic: they won’t go back to full openness, but they also won’t maintain a closed model that limits the ecosystem.

For Meta, this is crucial because their strength lies in their massive active user base. If they manage to integrate these models into Instagram, WhatsApp, and Messenger for intelligent content generation and personalized recommendations, they could create network effects across their entire infrastructure. The AI market is clearly segmented: premium enterprise vs. mass consumer. Meta is playing in the second field.

Regarding investments, Meta has been spending hundreds of billions on AI, which puts pressure on margins. Shares have recently fluctuated around $570-$580. Investors are watching closely: they need to see clear monetization pathways. If Alexandr Wang’s strategy works, it could open new subscription options or significantly improve advertising accuracy.

The risk remains the performance gap compared to established competitors and security risk management. But honestly, Meta is charting a unique path in the AI race. The bet on consumer scale instead of enterprise premium is risky, but if it works, it could completely change the game.