In a post on Facebook, Zuckerberg said Meta was building a new data center “so large it would cover a significant part of Manhattan.”

Zuckerberg said he expected Meta AI to become “the leading assistant serving more than 1 billion people” and was building an “AI engineer” that would start contributing code to the company’s research and development efforts.

The company is investing between $60 billion and $65 billion in 2025 to help build out its AI infrastructure, a significant increase from the $38 billion to $40 billion Meta spent in 2024.

Zuckerberg also said Meta planned to end the year with more than 1.3 million GPUs, a significant increase on the company’s 2024 goals.

The chips, which are vital for training and powering AI systems, have been in short supply over the past two years as demand for them spiked. Big Tech companies have been fighting to stockpile some of the most powerful chips.

“This is a massive effort, and over the coming years it will drive our core products and business, unlock historic innovation, and extend American technology leadership. Let’s go build,” Zuckerberg wrote in the post.

Meta’s announcement comes just days after the company’s AI rival, OpenAI, announced a $500 billion AI infrastructure project backed by SoftBank and Oracle, among others.

The Stargate Project, announced by Donald Trump on Monday, plans to build out AI data centers in the U.S.

Representatives for Meta did not immediately respond to Fortune’s request for more details on its plans.

AI vs. coding jobs

It’s not the first time Zuckerberg has suggested using AI to automate coding tasks internally.

In an interview on The Joe Rogan Experience earlier this month, Zuckerberg said he expected AI would soon be able to do the job of a “mid-level engineer” at Meta.

“Probably in 2025, we at Meta, as well as the other companies that are basically working on this, are going to have an AI that can effectively be a sort of mid-level engineer that you have at your company that can write code,” Zuckerberg said in the interview.

“In the beginning, it’ll be really expensive to run, and you can get it to be more efficient,” he added. “And over time it’ll get to the point where a lot of the code in our apps and including the AI that we generate is actually going to be built by AI engineers instead of people engineers.”

Meta’s Big Tech rival, Google, has also been experimenting with using AI to code internally.

Google CEO Sundar Pichai said last year that more than a quarter of new code created at Google was being generated by AI and overseen by human engineers.

The trend has caused alarm among industry employees, especially computer science graduates trying to enter the sector in junior roles.