Artificial Intelligence

Meta to Invest 'Hundreds of Billions' to Build Superintelligence

JUL 15 2025·Updated JUL 16 2025

Story

Above: Mark Zuckerberg and the Meta group's AI logo. Image credit: Vincent Feuray/Hans Lucas/AFP/Getty Images

The Facts

Meta CEO Mark Zuckerberg announced on Monday that he plans to invest "hundreds of billions of dollars" in artificial intelligence computing infrastructure. The company's first one-gigawatt data center mega-project — dubbed Prometheus after the Greek titan — is set to launch online in 2026.
Bloomberg.com
The Telegraph
Threads
A one-gigawatt data center consumes more power than about 750,000 standard homes. Zuckerberg said the company is building multiple multi-gigawatt clusters, including Hyperion, which will scale up to 5 gigawatts over several years and cover a significant portion of Manhattan's footprint.
The Telegraph
Hexatronicdatacenter
Facebook
Meta has launched Meta Superintelligence Labs, led by Alexandr Wang, following a $14.3 billion investment in Scale AI, to focus on developing artificial superintelligence that exceeds human intelligence capabilities.
Yahoo! News
The New York Times
The company has embarked on an aggressive hiring campaign offering packages as high as $300 million over four years to recruit top AI researchers from competitors, including OpenAI, Google DeepMind and Anthropic.
Fortune India
Datacentremagazine
Meta's superintelligence lab has reportedly discussed potentially abandoning its open-source AI model approach in favor of developing closed models.
New York Post
The New York Times
Meta's advertising business is funding the infrastructure investment. Zuckerberg stated that the company has the capital from its core business operations to support the superintelligence effort.
DIGITIMES
TechCrunch
PYMNTS

Sources Split

LEFT

CENTER

RIGHT

ESTAB
CRIT

CENTER

PRO
ESTAB

The Spin

Narrative A

Meta's massive AI investment represents necessary catch-up spending after falling behind competitors like OpenAI and Google. The company's open-source approach hasn't delivered the breakthrough results needed, and Zuckerberg's aggressive hiring and infrastructure push show a serious commitment to regaining leadership in the AI race.

TechCrunch

Investopedia

The Times (UK)

Narrative B

This is another costly Meta moonshot that may not deliver returns, similar to the company's $60 billion metaverse spending since 2020, which has yielded no clear profits. The shift away from open-source models contradicts years of company philosophy, and the astronomical spending raises questions about the sustainability of its business strategy.

The New York Times

The Healthcare Innovation Letter

Techno-skeptic narrative

AI-driven data centers consume vast amounts of energy and water due to the high computational power required for training and operating large models. Cooling systems alone demand significant electricity and water to prevent overheating. As AI usage grows, so does its environmental footprint, and the need for more sustainable infrastructure and greener technologies increases — raising concerns about the toll of expanding digital infrastructure.

Environmental and Energy Study Institute

Algaesys

Techno-optimist narrative

AI-driven data centers are vital for powering advanced technologies that support healthcare, finance, research, and communication. They enable real-time data processing, large-scale model training, and global connectivity. As the backbone of AI innovation, these centers accelerate automation, enhance decision-making, and drive digital transformation across industries — making them essential for economic growth and technological advancement in the modern world.