Following its reorganization, Meta’s AI unit will speed up the creation and release of AI products and features from both Facebook, Instagram and WhatsApp. The company hopes to become more competitive in the fast-changing AI area by tackling players such as OpenAI, Google and Microsoft.
What Does Meta’s New AI Structure Look Like?
According to an internal note by Chris Cox, Meta’s AI workforce is now being organized into two main branches.
- AI Products Team: Connor Hayes is in charge of the AI Products Team which specializes in making and delivering products with AI elements.
- AGI Foundations Unit: The AGI Foundations Unit is headed by Ahmad Al-Dahle (chair) and Amir Frenkel, working to boost AGI research and development.
The goal with these changes is to increase the number of employees who own the workforce, along with clearly outlining how teams rely on one another, helping drive rapid change and adoption of AI technologies. It is worth mentioning that no executives are quitting, no employees are being let go and some managers are taking over from other departments.
What Produced Meta’s Choice to Restructure Its AI Teams?
The company shifting its AI teams is happening at a time when competition in AI is rising. Because of advances made by OpenAI and Google, Meta has responded by boosting its progress in artificial intelligence. The restructuring falls under a wider plan aimed at improving how the company handles day-to-day work and quickly develops innovations for users.
How it Affects Meta’s Responsible AI Work
While restructuring, Meta split up its Responsible AI (RAI) team. The majority of RAI employees are expected to join Meta’s generative AI unit which is developing AI products. Individuals from multiple areas will become a part of the AI infrastructure unit, charged with developing and running AI tools. While the move is designed to guide product development with responsible AI, many still wonder how Meta will oversee the safety and ethics of its AI going forward.
Why Meta is Building an AI Infrastructure
To drive its AI goals, Meta is enlarging its infrastructure. Nvidia’s H100 GPUs will account for around 350,000 by years end, besides those coming from other manufacturers, totaling about 600,000 GPUs. This substantial effort demonstrates Meta’s plan to make advanced AI systems that will aid many AI models and apps.
Looking Ahead
Moving AI teams into one unit reflects that Meta is now focusing more on developing and researching AI products. Meta wants to develop AI solutions more rapidly for its users by making its operations more efficient and concentrating on generative AI and AGI. As the business grows its AI technology, it must make sure to also focus on ethics, in order to responsibly deploy these technologies.
