Luma AI, a Silicon Valley startup specializing in multimodal artificial intelligence, is rapidly scaling its operations following a $900 million funding round led by Saudi AI firm HUMAIN. The company’s growth, from 30 employees in early 2024 to over 160 today, reflects a broader trend: AI development is increasingly capital-intensive and globally distributed.
Rapid Growth and Strategic Expansion
CEO Amit Jain emphasized that aggressive growth is essential for survival in the startup world. Luma has already established offices in London and Seattle, adding 20-25 employees monthly. This expansion coincides with the launch of key products like Dream Machine, a video generation platform that attracted one million users in just four days, and Ray3, the first reasoning video model.
Infrastructure: The Next Bottleneck for AI
To support its escalating computational needs, Luma is partnering with HUMAIN on “Project Halo,” a massive AI infrastructure initiative. This project aims to deliver up to two gigawatts of capacity by the early 2030s, collaborating with industry giants NVIDIA and AMD. Jain frames this as one of the largest infrastructure build-ups in the AI sector, rivaling the scale of leading AI labs. The race for compute power is now a defining factor in AI dominance.
Filling a Cultural Gap in Generative AI
Luma is also strategically expanding into the Middle East to address a critical gap in generative AI: the underrepresentation of diverse cultures. Jain highlights that AI models are only as good as the data they’re trained on, and current datasets lack sufficient Arabic representation. The company is collaborating with Saudi Arabian partners to build the world’s first Arabic world model.
The Future of AI Compute: A Shift to the Middle East?
Jain predicts energy and computational capacity will become the primary bottlenecks for AI development. The Middle East, with its access to land, energy, and capital, is uniquely positioned to become a major exporter of AI compute. This could reshape global digital infrastructure in a way that mirrors oil’s historical influence on the global economy.
“History of our time won’t be in archaeology,” Jain warned. “It will be on the internet.”
The company’s rapid expansion and massive infrastructure investments underscore a pivotal moment in the AI landscape. If the Middle East leverages its advantages, it could emerge as a dominant force in shaping the future of artificial intelligence.
