Skip to content

OGAS: The Soviet Internet That Never Was

Abstract

In 1962 (seven years before the first ARPANET message) the Soviet cybernetician Viktor Glushkov proposed OGAS, the All-State Automated System: a three-tier computer network linking a central hub in Moscow to some 200 regional centers and up to 20,000 local terminals in factories across the USSR, designed to run the entire planned economy in real time. The technology was plausible; the mathematics was sound; the state that would have benefited most refused to build it. OGAS died not of engineering failure but of bureaucratic self-defense, ministries and managers who understood precisely what a transparent information system would do to their power. It is the great counterfactual of networking history: the country that most needed a national computer network was institutionally incapable of wanting one.

Cybernetics Rehabilitated

Under Stalin, cybernetics had been denounced as a “bourgeois pseudoscience.” After his death the Soviet Union reversed itself with remarkable speed: the science of feedback and control, it turned out, looked like exactly what a centrally planned economy needed. By the late 1950s the USSR had world-class computing hardware for the era (Sergei Lebedev’s MESM and BESM machines) and a rising school of mathematicians who believed the plan could be computed.

The first to say so concretely was the military computer scientist Anatoly Kitov, who in 1959 wrote directly to Khrushchev proposing that the military’s computers be used at night to manage the civilian economy, a unified national network of computing centers. The military took offense at sharing its machines with civilian planners; Kitov was expelled from the party and dismissed. The pattern (a rational proposal destroyed by the institution it threatened) would repeat.

Glushkov’s Vision

Viktor Glushkov, director of the Institute of Cybernetics in Kyiv (where Kateryna Yushchenko headed programming theory), inherited the idea and enlarged it. His 1962 proposal, the Obshchegosudarstvennaya avtomatizirovannaya sistema (All-State Automated System for the Gathering and Processing of Information), described a hierarchical network: one central computing center in Moscow, up to 200 mid-level centers in major cities, and up to 20,000 terminals in enterprises, communicating in real time over the existing telephone infrastructure, with any authorized user able to reach any other node.

Glushkov’s ambitions went beyond data collection. He envisioned electronic payments replacing cash, mathematical optimization of the plan itself, and what he called a “thinking envelope” for the economy. He estimated the project would take decades and cost more than the space and nuclear programs combined (by one estimate as much as 100 billion rubles) and projected it would repay itself several times over within fifteen years. For a state whose official ideology was scientific planning, it was the logical endpoint of the whole Soviet project.

October 1, 1970: Death by Committee

OGAS spent the 1960s in study commissions while Glushkov lobbied relentlessly. The decisive meeting came before the Politburo on October 1, 1970. Glushkov made his case. The Minister of Finance, Vasily Garbuzov, made the counter-case, reportedly arguing that computers were machines for lighting chicken coops, and that his ministry’s existing accounting methods sufficed. Garbuzov understood the stakes perfectly: a network that made economic information instantly visible in Moscow would make the Ministry of Finance’s role as gatekeeper of that information redundant. Factory directors quietly agreed, for their own reason, real-time reporting would expose the padded quotas and informal barter that made the planned economy actually function.

The Politburo declined to fund the full system. The 24th Party Congress in 1971 endorsed only “local information management systems”, thousands of disconnected factory-level computer installations, which were duly built through the 1970s and never linked into a network. Glushkov kept fighting for the larger vision until his death in 1982, at 58. By then the Soviet Union had also abandoned indigenous computer architecture in favor of cloning IBM’s System/360 (the ES EVM decision) and the moment had passed.

Info

The historian Benjamin Peters, in the definitive study How Not to Network a Nation, distilled the irony: the capitalist Americans built ARPANET by behaving like socialists (collaborative, state-funded, non-commercial) while the socialist Soviets failed to network their nation because their institutions behaved like competing capitalists, each ministry and factory defending its private informational fiefdom.

Dead End: The Hardware Was Never the Problem

OGAS is the cleanest historical demonstration that networks fail on institutional, not technical, grounds. Nothing in the design was beyond Soviet capability: the USSR had the mathematicians, adequate machines, and a telephone network. What it did not have was a single institution that would gain from transparency. Information in the Soviet economy was not a resource to be optimized; it was the currency of bureaucratic power, and OGAS proposed to devalue it. Every actor whose cooperation was required (finance ministry, statistical agency, factory directors) correctly perceived the system as an audit of themselves.

The comparison with its Western contemporary is exact and damning. ARPANET (first link, 1969) succeeded as a decentralized network among institutions that trusted each other and gained from sharing. OGAS proposed a decentralized architecture for institutions locked in zero-sum competition, under an ideology that could not admit this was so. Chile’s Project Cybersyn (the other great socialist network experiment) at least got built before politics destroyed it; OGAS never escaped the committee room. The Soviet Union thus entered the 1980s with tens of thousands of computers and no network worth the name, and when the internet finally arrived in Russia, it arrived as an import.

📚 Sources