Recently, drone strikes linked to the ongoing conflict targeted key cloud infrastructure in the Gulf, causing multi-billion-dollar AI data centers to go offline within minutes. Businesses rushed to reroute their workloads, revealing a clear and troubling fact: the region’s centralized AI infrastructure is just as vulnerable as any physical pipeline.

Now, imagine a different architecture. A global network of GPUs scattered across homes, studios, and offices, all connected in a single distributed grid. There is no single point of failure, and no geopolitical event can bring it all down. “It is global. You [would] need to nuke the whole planet to shut it down,” says Ilman Shazhaev, founder and CEO of Dizzaract, a global game development studio and the group behind FAR Labs, a new AI infrastructure venture.

FROM CENTRALIZATION TO DISTRIBUTED RESILIENCE

Shazhaev’s model offers a new way to think about the AI economy. Rather than building huge, centralized data centers like the UAE’s National Program for Artificial Intelligence or Saudi Arabia’s Data and Artificial Intelligence Authority, FAR Labs links unused GPUs in homes and studios into a distributed, blockchain-like network.

“In Bitcoin, you have many nodes processing transactions. Even when it has no utility as money, it serves as digital gold. Everyone just buys it and holds it. But why? Because its architecture is super secure. In 17 to 18 years, no one has been able to hack it,” says Shazhaev. FAR Labs applies this philosophy to AI compute. “If the war happened… has [it had] any effect on operating the Bitcoin blockchain? No. Nothing. So for us, it’s similar… a balanced, redistributed, [and] mathematically optimized framework.”

This architecture solves two of AI’s biggest headaches: cost and latency. Training models is expensive, but inference, the real-time processing of AI requests, is where hyperscalers bleed billions. “Inference is a major cost… with such a solution, we can make it officially the cheapest inference in the market right now,” says Shazhaev.

GAMING AS THE ULTIMATE STRESS TEST

FAR Labs’ background in gaming influenced how the platform was built. Gaming needs a lot of computing power and quick responses in detailed 3D worlds. “If text prompts are so expensive… imagine you have high-level 3D visuals. We found it was almost impossible. The most efficient solution was to handle processes locally, within user hardware,” Shazhaev explains.

This approach turns unused GPUs into assets that can earn money. “Every home can become a small data center. You can rent out your power to the common network, or use it for your own needs.” In a region where 93% of the population engages in gaming and e-sports, the potential is massive. FAR Labs has already registered over 1,200 node operators even before public onboarding.

DEMOCRATIZING AI INFRASTRUCTURE

The company’s approach is all about ownership. “We are not competing with big tech. We are optimizing the hardware they have already sold,” Shazhaev says. Users provide compute, and the platform rewards them based strictly on what they contribute: “Just compute delivered, rewards earned.”

This model could work alongside the Gulf’s centralized AI plans rather than compete with them. “You can have your own centralized big data centers, and at the same time, you can have distributed ones. It’s just about playing smart,” says Shazhaev. The system also protects security and privacy: user data stays on the platform, and only encrypted prompts are handled by nodes.

Distributed networks are inherently resilient. If a regional data center shuts down, workloads can quickly move to nodes around the world. “Within our launch… we will be talking with businesses… Once the network is live, all [compute] is happening within hours, if not minutes. It’s user-friendly, a very fast switch,” he says.

Shazhaev’s comparison to Bitcoin highlights the main advantage: decentralization creates an infrastructure that is almost impossible to take down with local attacks. 

The Gulf’s strong gaming culture boosts FAR Labs’ potential. The network can connect not just PCs but also Mac Minis, PlayStation consoles, and other powerful devices. “It’s not only about monetization of your hardware; it’s a depreciating asset that can now generate passive income,” Shazhaev says.

This approach offers a bold but practical vision: a locally owned, distributed, and resilient AI infrastructure that supports national AI investments and builds a truly participatory economy.