Who Owns The Future Of Compute? The Quest To Make AI Open For All

The Future of Artificial Intelligence: A Battle for Computing Power

The world of artificial intelligence (AI) is at a crossroads. The next decade of innovation will be shaped by the availability of computing power, and the companies that control this resource will hold significant sway over the development of AI. The fight for computing power is no longer just about algorithms; it’s about access to the hardware that powers AI systems. Graphics processing units (GPUs), in particular, have become the most sought-after resource in the digital economy.

The demand for GPUs is skyrocketing, with orders for Nvidia’s newest Blackwell GPUs exceeding 3.6 million units, driven largely by major cloud service providers. This has created a challenging environment for smaller startups and public institutions, which are struggling to access the computing power they need to develop and train AI models. The concentration of computing power in the hands of a few large corporations has become a major economic and political issue, with significant implications for the future of AI.

The Dominance of Nvidia and the Case for Open Infrastructure

Nvidia’s dominance of the global GPU market is unparalleled, with the company controlling around 94% of the market. This has created a situation where a handful of large corporations, including Nvidia, decide who gets access to GPUs, how much they cost, and which projects move forward. The consequences of this concentration of power are far-reaching, shaping who gets to innovate, how fast costs fall, and which countries hold leverage in the coming AI economy.

In response to this challenge, a growing number of researchers and builders are rethinking how computing power should be owned and shared. David and Daniil Liberman, founders of Gonka, are developing a community-governed network that allows participants to contribute and exchange computing power. Their goal is to make AI access as open as the early internet, rather than a gated service run by a few dominant players. The idea of compute as shared infrastructure is gaining momentum, with many experts drawing inspiration from decentralized systems like Bitcoin.

The Paradox of Distributed Systems

Decentralized networks like Gonka offer a promising solution to the problem of concentrated computing power. However, they also present significant challenges, including the need for rigorous verification and incentives aligned with performance. Without these safeguards, community-run networks risk slipping into inefficiency or manipulation. The paradox of distributed systems is that the more open they become, the more coordination they demand. Researchers at EPOCH AI, a nonprofit research group, have highlighted the need for careful design and governance to ensure that decentralized networks can scale without reproducing the same bottlenecks they hope to eliminate.

History has shown that decentralized systems can slowly recentralize, as power tends to pool where capital and capacity accumulate. To avoid this fate, the Libermans have intentionally avoided features like delegation in the Gonka protocol, which could give advantages to large players. Instead, control of the network has shifted to its users, ensuring that no single entity can unilaterally decide how the network evolves.

Sovereignty and the New Power Map

The battle for computing power has moved beyond the technical realm and into the world of geopolitics. Governments in Europe, Asia, and Africa are now linking AI sovereignty with control over computing resources. The concern is not about control itself, but about the dominance of the US and China, which can effectively cut them off from the prosperity that AI will bring. Decentralization is seen as the only way to ensure that citizens have equal access to the full benefits of AI.

Around the world, policymakers are beginning to treat compute as critical infrastructure, not just an industrial asset. Regional alliances are building shared digital and compute hubs to strengthen local capacity and reduce dependence on US and Chinese providers. The World Economic Forum has noted that access to computing power now influences productivity and economic inclusion, with nearly half of US workers using AI tools daily. The future of AI will be shaped by the ability of governments and companies to balance the need for computing power with the need for sovereignty and control.