Davos: G42 CEO Peng Xiao on UAE Chip Imports, AI Developments

Artificial Intelligence as a National Priority: A Discussion with the UAE Representative

Key Concepts:

• AI as a Physical Utility: The shift in perception of AI from a conceptual tool to a fundamental infrastructure component.
• AI Sovereignty: The importance of national control and independence over AI infrastructure.
• Stargate: The large-scale AI campus being developed in Abu Dhabi, powered by significant energy resources.
• Inferencing vs. Training: The distinction between the computational demands of developing AI models (training) and deploying them for use (inferencing).
• Intelligence Grids: The UAE’s vision for a globally connected AI network serving a large underserved population.
• PAX Silicon Framework: A US initiative focused on securing critical materials and semiconductors supply chains.

I. The Global Imperative of AI Investment

The speaker emphasizes that AI is no longer a future consideration but a present-day necessity. He notes a significant shift in governmental perspectives over the past two years, with AI now being recognized as a “real almost physical facility, physical utility” that nations are actively consuming. Governments are focused on two key priorities: ensuring AI deployment aligns with national values and culture, and establishing infrastructure independence to prevent external disruption. He argues that failing to prioritize AI investment carries significant risks, potentially leading to national disadvantage in governance, economic competitiveness, and security.

II. The UAE’s Leadership in AI Development

The UAE positions itself as a leader in the AI space, citing initiatives predating the recent surge in global interest. In 2018, the UAE launched the Embassy University for AI and established a federal-level Ministry of AI. This proactive approach stemmed from a foresight that AI would become indispensable. The UAE’s strategy centers on international partnerships, particularly with US companies like Microsoft and OpenAI, recognizing that AI development requires global collaboration.

A pivotal moment was the announcement in May, during President Trump’s visit, of the “Stargate” project – a massive AI campus powered by up to five gigawatts of dedicated energy resources. This demonstrates a substantial commitment to AI infrastructure. As the speaker stated, “Abu Dhabi was already leading A.I. before it was cool.”

III. Stargate Project: Progress and Infrastructure Details

The Stargate project is currently under rapid construction. Over 7,000 workers and 100 cranes are operating 24/7. The first 200 megawatts of capacity are expected to come online within the next few months, with plans to add 200-500 megawatts per quarter, ultimately reaching the full five-gigawatt capacity.

Securing access to advanced AI chips was a critical step. The UAE successfully obtained export licenses for Nvidia chips, as well as chips from Cerebras and EMG. This required providing comprehensive security assurances to the US, demonstrating the UAE’s ability to safeguard sensitive defense technologies and prevent unauthorized access or transfer. The speaker clarified that these assurances were not merely theoretical pledges but were backed by three years of demonstrated security practices.

IV. Addressing Concerns Regarding AI Investment & ROI

The speaker acknowledges concerns about the substantial capital expenditure (CapEx) in the AI sector and the potential difficulty in generating a return on investment. He identifies three primary anxieties: job security due to AI automation, the potential strain on energy resources, and the risk of falling behind in the global AI race.

However, he argues against pausing investment, stating that a slowdown by the US and its allies would allow other nations to gain a competitive advantage and lead to widespread AI diffusion without the influence of Western values and technology. He advocates for “doubling down” on AI development, even in the face of these challenges, emphasizing the UAE’s reliable and affordable energy supply and its strategic connectivity.

V. Infrastructure Demand, Supply Chain & the PAX Silicon Framework

The speaker believes that demand for AI infrastructure currently exceeds supply, particularly in the area of inferencing – the process of deploying AI models for real-world applications. While a limited number of companies will continue to dominate AI training (model development), the need for inferencing capacity is rapidly growing.

The UAE’s recent participation in the PAX Silicon framework with the US is seen as a significant opportunity for G42, the UAE’s AI company. This framework focuses on securing critical materials and semiconductors, and the speaker anticipates it will enable G42 to expand its role beyond infrastructure development to become a key enabler in the global AI supply chain. Jacob Helbach, a key figure in the framework, is seen as a valuable partner.

VI. G42’s Global Expansion & the “Intelligence Grids” Vision

The UAE’s strategic advantages – abundant energy and exceptional network connectivity – are central to its global expansion plans. The country can serve approximately 4 billion people with low latency, making it a crucial node in a future “Intelligence Grids” network. This network aims to deliver AI services to a vast, underserved global population.

The partnership with Microsoft is critical, enabling Microsoft to serve its customers directly from the UAE’s infrastructure. The Kenya project is progressing, despite facing typical challenges associated with frontier markets, particularly regarding commercialization and generating near-term returns.

VII. Regional Collaboration & a Collective Future

The speaker views AI ambitions in neighboring countries like Qatar and Saudi Arabia not as competition, but as a positive regional trend. He credits Abu Dhabi with initiating this movement and believes that the collective resources and energy of the region should be leveraged to benefit humanity as a whole. He stated, “I actually don’t see any competition. I think it’s wonderful.”

Technical Terms & Concepts:

• Inferencing: The process of using a trained AI model to make predictions or decisions on new data.
• Training: The process of developing an AI model by feeding it large amounts of data.
• Latency: The delay between a request and a response in a network.
• Gigawatt (GW): A unit of power equal to one billion watts.
• Megawatt (MW): A unit of power equal to one million watts.
• PAX Silicon Framework: A US-led initiative to bolster the supply chain for critical materials and semiconductors.

Conclusion:

The interview highlights the UAE’s proactive and ambitious approach to AI development, positioning itself as a global leader through substantial investment in infrastructure, strategic partnerships, and a long-term vision for a connected AI future. The speaker underscores the urgency of prioritizing AI as a national imperative, emphasizing the risks of falling behind and the potential benefits of leveraging AI for economic growth, national security, and global impact. The UAE’s “Intelligence Grids” concept represents a commitment to democratizing access to AI and serving a vast underserved population.