Shift in Open‑Source Strategy

Meta’s recent AI roadmap appears to be evolving. After launching the Llama 4 model as an open‑source offering, the company faced criticism for a lackluster release that included allegations of gaming AI benchmarks and a postponed "Behemoth" version. In response, Mark Zuckerberg authored a July 30th memo on "personal superintelligence" that emphasized the need for caution. He wrote that Meta must be "careful about what we choose to open source" to mitigate potential safety risks. This sentiment signals a possible departure from the open‑source philosophy that Zuckerberg previously described as "the path forward."

Avocado Model and Internal Reorganization

According to Bloomberg, Meta is now developing a new AI model internally code‑named Avocado. Unlike prior releases, the company may charge for access to this model, suggesting a monetization strategy that contrasts with the free distribution of Llama. The shift aligns with internal changes highlighted by the New York Times, which reported that Zuckerberg is spending considerable time with a group of new hires known as TBD Lab. This team operates out of a "siloed space" near Zuckerberg’s office at Meta’s headquarters, indicating a focused effort on advanced AI development within a more controlled environment.

These developments collectively point to a strategic realignment at Meta, where safety considerations, commercial viability, and tighter internal coordination appear to be reshaping the company’s approach to artificial intelligence.

Usado: News Factory APP - descoberta e automação de notícias - ChatGPT para Empresas