Meta CEO Mark Zuckerberg has outlined the AI ambitions of the social media giant for 2025 and stated that the upcoming large language model (LLM) of the company, Llama 4, will be “State of the Art”. Meta AI is in particular one of the most used chatbots, thanks to the direct availability on all services of the company, including WhatsApp, Facebook, Instagram and even the smart glasses from Ray-Ban. Zuckerberg, however, strives that the chatbot will reach more than 1 billion people, a big leap compared to the 600 million monthly active users he had last year.
In subsequent reports on Threads, Zuckerberg wrote: “This will be a decisive year for AI. I expect Meta AI to be the leading assistant in 2025 who will serve more than 1 billion people, Llama 4 will become the leading state-of-the-art model, and that we will build an AI engineer that is increasingly larger quantities of code will contribute to our R&D. attempts. To make this possible, Meta is building a 2GW+ data center that is so large that it would cover a considerable part of Manhattan. ”
“We will bring the computing power online in '25 ~ 1GW and we will close the year with more than 1.3 million GPUs. This year we intend to invest $ 60-65 billion in Capex and at the same time grow our AI teams considerably, and we have the capital to keep investing in the coming years. “
The announcement of Zuckerberg to invest billions in AI infrastructure comes when Rival Microsoft announced an investment of $ 80 billion to build data centers, while Chatgpt-maker OpenAi has joined a joint venture called Stargate.
Zuckerberg strives for 'American technological leadership':
Zuckerberg added that the new investments will “unlock historical innovation and increase American technological leadership.”
He wrote: 'It is a huge effort, and in the coming years it will stimulate our core products and activities, unlock historical innovation and increase American technological leadership. Let's build! 💪 ”added Meta's CEO in the Threads post. “