Meta is turning ambitious claims about artificial intelligence into concrete actions. After announcing a large capital spending plan for AI last year, this technology group has officially launched an AI infrastructure initiative called Meta Compute, aimed at strengthening core capabilities for new generation AI models and products.
At the summer financial report meeting last year, Ms. Susan Li, Meta's Chief Financial Officer, emphasized that building a leading AI infrastructure will become a key competitive advantage.
According to her, infrastructure not only determines the ability to train models but also directly affects the product experience that users access.
To date, Meta shows that they are fulfilling that commitment. In a post on Threads on Monday (January 12) local time, CEO Mark Zuckerberg said the company plans to strongly expand energy consumption to serve AI.
Meta is planning to build dozens of gigawatts in this decade, and hundreds of gigawatts or more in the future," Mark Zuckerberg wrote, while affirming that the way of designing, investing and cooperating in infrastructure construction will become a long-term strategic advantage.
Gigawatt is a unit of power equivalent to one billion watts, showing the enormous scale of this plan. AI, especially generative AI, is becoming a major energy-consuming industry.
Some estimates suggest that electricity demand for AI in the US may increase from about 5 gigawatts currently to 50 gigawatts in the next decade, creating unprecedented pressure on electricity systems and data infrastructure.
To lead Meta Compute, Zuckerberg announced three key leaders. First is Santosh Janardhan, head of Meta's global infrastructure division and has been with the company since 2009.
He will be in charge of engineering architecture, software systems, silicon programs, developer productivity, as well as building and operating a global data center network.
The second person is Daniel Gross, who joined Meta last year and is the co-founder of Safe Superintelligence with former OpenAI science director Ilya Sutskever.
Gross will lead the new group to be responsible for long-term capacity strategy, supplier partnerships, industry analysis, planning and business models.
Finally, Dina Powell McCormick, a former US government official, currently holds the role of Chairman and Vice Chairman of Meta's Board of Directors.
She will be in charge of cooperation with governments to support the construction, deployment, investment and funding of the corporation's AI infrastructure.
The Meta Compute initiative takes place in the context of the AI infrastructure construction race heating up.
Microsoft is accelerating cooperation with AI infrastructure providers, while Alphabet (Google's parent company) has acquired the data center enterprise Intersect.
