Zuckerberg's "sky-high poaching" begins to pay off as Meta (META.US) reveals that a new AI team has delivered core models this month.

Meta's Chief Technology Officer announced on Wednesday that the company's newly established artificial intelligence laboratory has delivered the first batch of highly anticipated AI models internally this month. During a press briefing at the World Economic Forum in Davos, Chief Technology Officer Andrew Bosworth stated that the models developed by Meta's Super Intelligence Lab team, formed last year, have shown immense potential. Bosworth said, "They've been at this for basically only six months, if not less," and added that the team's AI models have performed "remarkably well." Media reports in December stated that Meta is developing a text AI model codenamed Avocado, set to be released in the first quarter, and also working on an image and video model codenamed Mango. Bosworth did not specify which models were delivered internally this time. Meta's efforts have been closely watched since CEO Mark Zuckerberg made significant moves to restructure the AI leadership, establish new labs, and poach talent with high salaries. Zuckerberg hopes the company can win in the fiercely competitive field of cutting-edge technology. Previously, Meta faced criticism for its performance of the Llama 4 model. Meanwhile, competitors like Alphabet Inc. Class C have gained an advantage in this transformative and lucrative AI competition. Bosworth noted that the technology is not yet complete. Regarding the development cycle in general, Bosworth said, "In the field of AI, there is still a massive amount of work to be done post-training to ensure that the delivered models are usable both internally and for consumers." However, he stated that Meta's major bet for 2025 is already showing promising returns.