‘Your AI Policy Is A Request, Not A Strategy’: Gonka’s Liberman Brothers On GPU Feudalism, The Energy Bottleneck, And Who Controls The Intelligence Layer

The AI race is often framed as a contest between fashions, chips, and capital. But a rising variety of voices argue the true chokepoint is easier and extra bodily: electrical energy, grid entry, and the handful of companies that management the infrastructure every thing else runs on. If you don’t personal compute, the argument goes, your AI technique is only a request.

The penalties of that logic are already enjoying out. Grid operators are sounding alarms, utility payments are climbing, and the query of who will get to entry AI — not simply use it, however run it, personal it, construct on it — is quietly turning into one of many extra consequential infrastructure questions of the last decade. For many of the world, the reply proper now’s: on another person’s phrases.

MPost sat down with Daniil and David Liberman, co-founders of Gonka, a decentralized AI compute community, to search out out whether or not there may be one other manner.

At what level did the AI vitality drawback change into an emergency? What breaks subsequent?

The emergency began when the limiting issue stopped being chips and have become electrical energy. For years, the AI business handled knowledge facilities like monetary objects. You elevate cash, purchase GPUs, and construct a large campus. But knowledge facilities are bodily objects tied to transformers, energy vegetation, and native grids. Those issues can not scale on the pace of software program. The latest July orders present that grids like PJM are actively managing round knowledge heart demand, utilizing backup era as an absolute final resort. What breaks subsequent is your entire financial mannequin. You can have billions of {dollars} in GPUs and actually nowhere to plug them in. Then the political layer breaks as households begin asking why their utility payments are funding an AI facility. The deeper mistake is assuming each AI computation requires a gigawatt scale campus. We want to maneuver compute towards vitality: a decentralized method routes workloads to current machines and areas the place energy is already accessible.

Who are the lords and serfs of GPU feudalism? What does the hire seem like?

The lords are the few actors who management the bodily and contractual layers, like superior chips, hyperscale clouds, and the vitality required to run them. The serfs are nearly everybody else. They are the startups, builders, and full international locations that should import intelligence as a service. This is strictly what we name GPU feudalism, a future the place folks change into tenants on another person’s compute property. The hire they pay is partly monetary, however the extra harmful hire is permission. A supplier can change your quota, droop your account, or silently swap the mannequin you’re utilizing. If 5 corporations and two states management the infrastructure, you get eight billion tenants. AI is pure data and may be copied nearly with out restrict. The mannequin suits on a flash drive, however the monopoly lives within the knowledge heart. That bodily bottleneck is strictly the place synthetic shortage is imposed on one thing meant to create international abundance.

Why does infrastructure possession decide what AI will get constructed?

An thought with out compute just isn’t an organization, only a doc. When compute entry is pricey or politically restricted, builders bodily change what they construct. A startup would possibly abandon an open supply mannequin just because a large cloud supplier subsidizes entry to a proprietary API. The infrastructure house owners dictate which {hardware} is supported, which areas get capability, and who will get precedence throughout shortages. This is why open weights are by no means sufficient. Publishing mannequin weights with out inexpensive execution is like sharing blueprints for an plane when one single company owns each runway. You additionally want verifiability. A decentralized system should show that the anticipated mannequin was really executed with out silent reductions in high quality. An agent that strikes cash wants greater than belief, it wants proof. If you don’t management compute, your AI coverage is a request somewhat than a method.

How shut are we to selecting between permissions somewhat than fashions?

We are already dwelling within the first model of that world. Users assume they’re selecting amongst dozens of AI merchandise, however many are simply interfaces constructed on the very same small group of mannequin suppliers. Today, permissions seem like geographic restrictions, utilization tiers, and API insurance policies. Tomorrow, when AI techniques change into important for financial productiveness, these restrictions change into structural. Imagine one engineer gaining access to the most effective coding agent whereas one other doesn’t. The firm or nation denied entry instantly falls behind. The final end result is a hierarchy of intelligence the place entry relies upon totally in your jurisdiction, employer, or id. People will technically be free to reject these techniques, however they’ll now not be economically aggressive with out them. We can nonetheless reverse this, however the window is measured in years. Every API shutdown or entry denial is simply one other commercial for decentralized AI.

Who wins as inference turns into the dominant battleground?

Training fashions generates the most important headlines, however inference is the place AI turns into an actual economic system. A mannequin is educated periodically, however inference occurs repeatedly each single time a consumer asks a query or a robotic comes to a decision. This shifts the aggressive panorama. While coaching requires huge and tightly related clusters, inference can ceaselessly be processed independently. You don’t want each request to run inside a single gigawatt campus. The winners will probably be techniques that scale back the price per helpful token, preserve high {hardware} utilization, and assure mannequin integrity. The losers will probably be companies completely counting on sponsored API pricing and international locations importing each unit of intelligence. This is the precise opening for distributed techniques. Decentralized compute can make the most of international provide and native vitality with out requiring one company to personal each facility.

What do 100 million requests throughout 26 international locations reveal about untapped compute?

They show that geographically distributed AI compute is now not only a theoretical idea. Before working an actual community, folks assumed the world contained a homogeneous pool of idle GPUs ready to be activated. The actuality is that {hardware}, networking, and availability differ drastically. The actually scarce useful resource is coordination. A community should uncover capability, measure efficiency, confirm execution, and create incentives that maintain operators on-line. This is why Gonka makes use of a transformer-based Proof of Work and Sprint mechanisms to confirm actual AI {hardware}. The untapped provide is huge, together with enterprise reserves, regional clouds, and college {hardware} that merely lacked a connection to international AI demand. Processing these workloads reveals that builders can devour decentralized inference by way of a standard interface while not having all compute to reside inside a company perimeter.

Can an open various nonetheless emerge, and at what scale?

Yes, nevertheless it won’t emerge mechanically from the present market leaders. An open various wants to achieve huge bodily scale to change into a real structural possibility. We are speaking about gigawatts of capability and tens of millions of superior GPUs. Bitcoin supplies the right lesson right here. It didn’t elevate billions of {dollars} to construct one monumental knowledge heart. Instead, it created an financial protocol that mobilized international bodily infrastructure. AI wants a comparable normal for helpful computation. Open fashions require open execution infrastructure. Economic rewards should circulation on to the individuals who present and enhance {hardware}. If AI will get a compute-aligned protocol, {hardware} innovation will transfer from gradual company roadmaps to international competitors.

Is shared compute a path to AI sovereignty for rising areas?

It completely may be, offered sovereignty is outlined appropriately. AI sovereignty doesn’t imply each nation should fabricate its personal chips and construct a home hyperscale cloud. Let’s be real looking — it’s not doable. A extra sensible definition is {that a} nation can’t be unilaterally disconnected from intelligence. Renting an API from a overseas firm is pure dependence. A nation with out compute will negotiate with AI precisely like a tenant negotiates with a landlord. Connecting home vitality, engineers, and knowledge facilities to an open community is true participation. No medium sized nation can outspend the United States or China individually. But if 200 international locations every contribute ten thousand GPUs on common, they create a shared community of two million GPUs. This coordination ensures sovereign entry to the intelligence layer, giving rising areas in Asia, Africa, and the Middle East an actual exit possibility from centralized management.

Is decentralized AI structurally aggressive, or solely filling Big Tech’s gaps?

Today, it’s doing each. Most decentralized networks begin in areas the place centralized suppliers are costly, restrictive, or just uninterested. That is totally regular. However, it turns into structurally aggressive when it creates an open normal that coordinates international vitality, rewards higher {hardware}, and provides builders a reputable exit technique. The related metric is rarely the market capitalization of a token. The decisive metric is beneficial AI computation processed per day. Decentralized techniques don’t want centralized suppliers to vanish. The web didn’t remove personal networks, and Bitcoin didn’t remove banks. Once builders have a viable various, incumbents are compelled to compete way more significantly on value, entry, and transparency.

Why would decentralized AI keep away from reconsolidation?

Reconsolidation is an actual danger. Decentralization just isn’t a everlasting state simply because somebody wrote the phrase in a white paper. AI compute has huge economies of scale, that means bigger operators can negotiate higher electrical energy contracts and purchase {hardware} extra cheaply. A permissionless community may simply consolidate round a couple of industrial operators. This is strictly why Gonka is constructed on compute-weighted governance, making certain that affect is related to verified computational contribution somewhat than easy token possession. The way forward for AI infrastructure should be ruled by the individuals who really convey intelligence into the community, not by idle capital. The objective is to not stop giant operators from current. Large operators are extremely environment friendly. But we’ve got to make sure they continue to be operators inside an open protocol somewhat than turning into absolute house owners of the protocol itself.

The submit ‘Your AI Policy Is A Request, Not A Strategy’: Gonka’s Liberman Brothers On GPU Feudalism, The Energy Bottleneck, And Who Controls The Intelligence Layer appeared first on Metaverse Post.