The Digital Antiquarian

The BESM-6

Seen from certain perspectives, Soviet computer hardware as an innovative force of its own peaked as early as 1968, the year the first BESM-6 computer was powered up. The ultimate evolution of the line of machines that had begun with Sergei Lebedev’s original MESM, the BESM-6 was the result of a self-conscious attempt on the part of Lebedev’s team at ITMVT to create a world-class supercomputer. By many measures, they succeeded. Despite still being based on transistors rather than the integrated circuits that were becoming more and more common in the West, the BESM-6’s performance was superior to all but the most powerful of its Western peers. The computers generally acknowledged as the fastest in the world at the time, a line of colossi built by Control Data in the United States, were just a little over twice as fast as the BESM-6, which had nothing whatsoever to fear from the likes of the average IBM mainframe. And in comparison to other Soviet computers, the BESM-6 was truly a monster, ten times as fast as anything the country had managed to produce before. In its way, the BESM-6 was as amazing an achievement on Lebedev’s part as had been the MESM almost two decades earlier. Using all home-grown technology, Lebedev and his people had created a computer almost any Western computer lab would have been proud to install.

At the same time, though, the Soviet computer industry’s greatest achievement to date was, almost paradoxically, symbolic of all its limitations. Sparing no expense nor effort to build the best computer they possibly could, Lebedev’s team had come close to but not exceeded the Western state of the art, which in the meantime continued marching inexorably forward. All the usual inefficiencies of the Soviet economy conspired to prevent the BESM-6 from becoming a true game changer rather than a showpiece. BESM-6s would trickle only slowly out of the factories; only about 350 of them would be built over the course of the next 20 years. They became useful tools for the most well-heeled laboratories and military bases, but there simply weren’t enough of them to implement even a fraction of the cybernetics dream.

A census taken in January of 1970 held that there were just 5500 computers operational in the Soviet Union, as compared with 62,500 in the United States and 24,000 in Western Europe. Even if one granted that the BESM-6 had taken strides toward solving the problem of quality, the problem of quantity had yet to be addressed. Advanced though the BESM-6 was in so many ways, for Soviet computing in general the same old story held sway. A Rand Corporation study from 1970 noted that “the Soviets are known to have designed micro-miniaturized circuits far more advanced than any observed in Soviet computers.” The Soviet theory of computing, in other words, continued to far outstrip the country’s ability to make practical use of it. “In the fundamental design of hardware and software the Russian computer art is as clever as that to be found anywhere in the world,” said an in-depth Scientific American report on the state of Soviet computing from the same year. “It is in the quality of production, not design, that the USSR is lagging.”

One way to build more computers more quickly, the Moscow bureaucrats concluded, was to share the burden among their partners (more accurately known to the rest of the world as their vassal states) in the Warsaw Pact. Several member states — notably East Germany, Czechoslovakia, and Hungary — had fairly advanced electronics industries whose capabilities in many areas exceeded that of the Soviets’ own, not least because their geographical locations left them relatively less isolated from the West. At the first conference of the International Center of Scientific and Technical Information in January of 1970, following at least two years of planning and negotiating, the Soviet Union signed an agreement with East Germany, Czechoslovakia, Bulgaria,  Hungary, Poland, and Romania to make the first full-fledged third-generation computer — one based on integrated circuits rather than transistors — to come out of Eastern Europe. The idea of dividing the labor of producing the new computer was taken very literally. In a testimony to the “from each according to his means” tenet of communism, Poland would make certain ancillary processors, tape readers, and printers; East Germany would make other peripherals; Hungary would make magnetic memories and some systems software; Czechoslovakia would make many of the integrated circuits; Romania and Bulgaria, the weakest sisters in terms of electronics, would make various mechanical and structural odds and ends; and the Soviet Union would design the machines, make the central processors, and be the final authority on the whole project, which was dubbed “Ryad,” a word meaning “row” or “series.”

The name was no accident. On the contrary, it was key to the nature of the computer — or, rather, computers — the Soviet Union and its partners were now planning to build. With the BESM-6 having demonstrated that purely home-grown technology could get their countries close to the Western state of the art but not beyond it, they would give up on trying to outdo the West. Instead they would take the West’s best, most proven designs and clone them, hoping to take advantage of the eye toward mass production that had been baked into them from the start. If all went well, 35,000 Ryad computers would be operational across the Warsaw Pact by 1980.

In a sense, the West had made it all too easy for them, given Project Ryad all too tempting a target for cloning. In 1964, in one of the most important developments in the history of computers, IBM had introduced a new line of mainframes called the System/360. The effect it had on the mainframe industry of the time was very similar to the one which the IBM PC would have on the young microcomputer industry 17 years later: it brought order and stability to what had been a confusion of incompatible machines. For the first time with the System/360, IBM created not just a single machine or even line of machines but an entire computing ecosystem built around hardware and software compatibility across a wide swathe of models. The effect this had on computing in the West is difficult to overstate. There was, for one thing, soon a large enough installed base of System/360 machines that companies could make a business out of developing software and selling it to others; this marked the start of the software industry as we’ve come to know it today. Indeed, our modern notion of computing platforms really begins with the System/360. Dag Spicer of the Computer History Museum calls it IBM’s Manhattan Project. Even at the time, IBM’s CEO Thomas Watson Jr. called it the most important product in his company’s already storied history, a distinction which is challenged today only by the IBM PC.

The System/360 ironically presaged the IBM PC in another respect: as a modular platform built around well-documented standards, it was practically crying out to be cloned by companies that might have trailed IBM in terms of blue-sky technical innovation, but who were more than capable of copying IBM’s existing technology and selling it at a cheaper price. Companies like Amdahl — probably the nearest equivalent to IBM’s later arch-antagonist Compaq in this case of parallel narratives — lived very well on mainframes compatible with those of IBM, machines which were often almost as good as IBM’s best but were always cheaper. None too pleased about this, IBM responded with various sometimes shady countermeasures which landed them in many years of court cases over alleged antitrust violations. (Yes, the histories of mainframe computing and PC computing really do run on weirdly similar tracks.)

If the System/360 from the standpoint of would-be Western cloners was an unlocked door waiting to be opened, from the standpoint of the Soviet Union, which had no rules for intellectual property whatsoever which applied to the West, the door was already flung wide. Thus, instead of continuing down the difficult road of designing its high-end computers from scratch, the Soviet Union decided to stroll on through.

An early propaganda shot shows a Ryad machine in action.

There’s much that could be said about what this decision symbolized for Soviet computing and, indeed, for Soviet society in general. For all the continuing economic frustrations lurking below the surface of the latest Pravda headlines, Khrushchev’s rule had been the high-water mark of Soviet achievement, when the likes of the Sputnik satellite and Yuri Gagarin’s flight into space had seemed to prove that communism really could go toe-to-toe with capitalism. But the failure to get to the Moon before the United States among other disappointments had taken much of the shine off that happy thought. ^[1]Some in the Soviet space program actually laid their failure to get to the Moon, perhaps a bit too conveniently, directly at the feet of the computer technology they were provided, noting that the lack of computers on the ground equal to those employed by NASA — which happened to be System/360s — had been a crippling disadvantage. Meanwhile the computers that went into space with the Soviets were bigger, heavier, and less capable than their American counterparts. In the rule of Leonid Brezhnev, which began with Khrushchev’s unceremonious toppling from power in October of 1964, the Soviet Union gradually descended into a lazy decrepitude that gave only the merest lip service to the old spirit of revolutionary communism. Corruption had always been a problem, but now, taking its cue from its new leader, the country became a blatant oligarchy. While Brezhnev and his cronies collected dachas and cars, their countryfolk at times literally starved. Perhaps the greatest indictment of the system Brezhnev perpetuated was the fact that by the 1970s the Soviet Union, in possession of more arable land than any nation on earth and with one of the sparsest populations of any nation in relation to its land mass, somehow still couldn’t feed itself, being forced to import millions upon millions of tons of wheat and other basic foodstuffs every year. Thus Brezhnev found himself in the painful position, all too familiar to totalitarian leaders, of being in some ways dependent on the good graces of the very nations he denigrated.

In the Soviet Union of Leonid Brezhnev, bold ideas like the dream of cybernetic communism fell decidedly out of fashion in favor of nursing along the status quo. Every five years, the Party Congress reauthorized ongoing research into what had become known as the “Statewide Automated Management System for Collection and Processing of Information for the Accounting, Planning, and Management of the National Economy” (whew!), but virtually nothing got done. The bureaucratic infighting that had always negated the perceived advantages of communism — as perceived optimistically by the Soviets, and with great fear by the West — was more pervasive than ever in these late years. “The Ministry of Metallurgy decides what to produce, and the Ministry of Supplies decides how to distribute it. Neither will yield its power to anyone,” said one official. Another official described each of the ministries as being like a separate government unto itself. Thus there might not be enough steel to make the tractors the country’s farmers needed to feed its people one year; the next, the steel might pile up to rust on railway sidings while the erstwhile tractor factories were busy making something else.

Amidst all the infighting, Project Ryad crept forward, behind schedule but doggedly determined. This new face of computing behind the Iron Curtain made its public bow at last in May of 1973, when six of the seven planned Ryad “Unified System” models were in attendance at the Exposition of Achievements of the National Economy in Moscow. All were largely hardware- and software-compatible with the IBM System/360 line. Even the operating systems that were run on the new machines were lightly modified copies of Western operating systems like IBM’s DOS/360. Project Ryad and its culture of copying would come to dominate Soviet computing during the remainder of the 1970s. A Rand Corporation intelligence report from 1978 noted that “by now almost everything offered by IBM to 360 installations has been acquired” by the Soviet Union.

Project Ryad even copied the white lab coats worn by the IBM “priesthood” (and gleefully scorned by the scruffier hackers who worked on the smaller but often more innovative machines produced by companies like DEC).

During the five years after the Ryad machines first appeared, IBM sold about 35,000 System/360 machines, while the Soviet Union and its partners managed to produce about 5000 Ryad machines. Still, compared to what the situation had been before, 5000 reasonably modern machines was real progress, even if the ongoing inefficiencies of the Eastern Bloc economies kept Project Ryad from ever reaching more than a third of its stated yearly production goals. (A telling sign of the ongoing disparities between West and East was the way that all Western estimates of future computer production tended to vastly underestimate the reality that actually arrived, while Eastern estimates did just the opposite.) If it didn’t exactly allow Eastern Europe to make strides toward any bold cybernetic future — on the contrary, the Warsaw Pact economies continued to limp along in as desultory a fashion as ever — Project Ryad did do much to keep its creator nations from sliding still further into economic dysfunction. Unsurprisingly, a Ryad-2 generation of computers was soon in the works, cloning the System/370, IBM’s anointed successor to the System/360 line. Other projects cloned the DEC PDP line of machines, smaller so-called “minicomputers” suitable for more modest — but, at least in the West, often more interesting and creative — tasks than the hulking mainframes of IBM. Soviet watcher Seymour Goodman summed up the current situation in an article for the journal World Politics in 1979:

The USSR has learned that the development of its national computing capabilities on the scale it desires cannot be achieved without a substantial involvement with the rest of the world’s computing community. Its considerable progress over the last decade has been characterized by a massive transfer of foreign computer technology. The Soviet computing industry is now much less isolated than it was during the 1960s, although its interfaces with the outside world are still narrowly defined. It would appear that the Soviets are reasonably content with the present “closer but still at a distance” relationship.

Reasonable contentment with the status quo would continue to be the Kremlin’s modus operandi in computing, as in most other things. The fiery rhetoric of the past had little relevance to the morally and economically bankrupt Soviet state of the 1970s and 1980s.

Even in this gray-toned atmosphere, however, the old Russian intellectual tradition remained. Many of the people designing and programming the nation’s computers barely paid attention to the constant bureaucratic turf wars. They’d never thought that much about philosophical abstractions like cybernetics, which had always been more a brainchild of the central planners and social theorists than the people making the Soviet Union’s extant computer infrastructure, such as it was, work. Like their counterparts in the West, Soviet hackers were more excited by a clever software algorithm or a neat hardware re-purposing than they were by high-flown social theory. Protected by the fact that the state so desperately needed their skills, they felt free at times to display an open contempt for the supposedly inviolate underpinnings of the Soviet Union. Pressed by his university’s dean to devote more time to the ideological studies that were required of every student, one young hacker said bluntly that “in the modern world, with its super-speedy tempo of life, time is too short to study even more necessary things” than Marxism.

Thus in the realm of pure computing theory, where advancement could still be made without the aid of cutting-edge technology, the Soviet Union occasionally made news on the world stage with work evincing all the originality that Project Ryad and its ilk so conspicuously lacked. In October of 1978, a quiet young researcher at the Moscow Computer Center of the Soviet Academy of Sciences named Leonid Genrikhovich Khachiyan submitted a paper to his superiors with the uninspiring — to non-mathematicians, anyway — title of “Polynomial Algorithms in Linear Programming.” Following its publication in the Soviet journal Reports of the Academy of Sciences, the paper spread like wildfire across the international community of mathematics and computer science, even garnering a write-up in the New York Times in November of 1979. (Such reports were always written in a certain tone of near-disbelief, of amazement that real thinking was going on in the Mirror World.) What Khachiyan’s paper actually said was almost impossible to clearly explain to people not steeped in theoretical mathematics, but the New York Times did state that it had the potential to “dramatically ease the solution of problems involving many variables that up to now have required impossibly large numbers of separate computer calculations,” with potential applications in fields as diverse as economic planning and code-breaking. In other words, Khachiyan’s new algorithms, which have indeed stood the test of time in many and diverse fields of practical application, can be seen as a direct response to the very lack of computing power with which Soviet researchers constantly had to contend. Sometimes less really could be more.

As Khachiyan’s discoveries were spreading across the world, the computer industries of the West were moving into their most world-shaking phase yet. A fourth generation of computers, defined by the placing of the “brain” of the machine, or central processing unit, all on a single chip, had arrived. Combined with a similar miniaturization of the other components that went into a computer, this advancement meant that people were able for the first time to buy these so-called “microcomputers” to use in their homes — to write letters, to write programs, to play games. Likewise, businesses could now think about placing a computer on every single desk. Still relatively unremarked by devotees of big-iron institutional computing as the 1970s expired, over the course of the 1980s and beyond the PC revolution would transform the face of business and entertainment, empowering millions of people in ways that had heretofore been unimaginable. How was the Soviet Union to respond to this?

Alexi Alexandrov, the president of the Moscow Academy of Sciences, responded with a rhetorical question: “Have [the Americans] forgotten that problems of no less complexity, such as the creation of the atomic bomb or space-rocket technology… [we] were able to solve ourselves without any help from abroad, and in a short time?” Even leaving aside the fact that the Soviet atomic bomb was itself built largely using stolen Western secrets, such words sounded like they heralded a new emphasis on original computer engineering, a return to the headier days of Khrushchev. In reality, though, the old ways were difficult to shake loose. The first Soviet microprocessor, the KP580BM80A of 1977, had its “inspiration” couched inside its very name: the Intel 8080, which was along with the Motorola 6800 one of the two chips that had launched the PC revolution in the West in 1974.

Yet in the era of the microchip the Soviet Union ran into problems continuing the old practices. While technical schematics for chips much newer and more advanced than the Intel 8080 were soon readily enough available, they were of limited use in Soviet factories, which lacked the equipment to stamp out the ever more miniaturized microchip designs coming out of Western companies like Intel.

One solution might be for the Soviets to hold their noses and outright buy the chip-fabricating equipment they needed from the West. In earlier decades, such deals had hardly been unknown, although they tended to be kept quiet by both parties for reasons of pride (on the Eastern side) and public relations (on the Western side). But, unfortunately for the Soviets, the West had finally woken up to the reality that microelectronics were as critical to a modern war machine as missiles and fighter planes. A popular story that circulated around Western intelligence circles for years involved Viktor Belenko, a Soviet pilot who went rogue, flying his state-of-the-art MiG-25 fighter jet to a Japanese airport and defecting there in 1976. When American engineers examined his MiG-25, they found a plane that was indeed a technological marvel in many respects, able to fly faster and higher than any Western fighter. Yet its electronics used unreliable vacuum tubes rather than transistors, much less integrated circuits — a crippling disadvantage on the field of battle. The contrast with the West, which had left the era of the vacuum tube behind almost two decades ago, was so extreme that there was some discussion of whether Belenko might be a double agent, his whole defection a Soviet plot to convince the West that they were absurdly far behind in terms of electronics technology. Sadly for the Soviets, the vacuum tubes weren’t the result of any elaborate KGB plot, but rather just a backward electronics industry.

In 1979, the Carter Administration began to take a harder line against the Soviet Union, pushing through Congress as part of the Export Administration Act a long list of restrictions on what sorts of even apparently non-military computer technology could legally be sold to the Eastern Bloc. Ronald Reagan then enforced and extended these restrictions upon becoming president in 1981, working with the rest of the West in what was known as the Coordination Committee on Export Controls, or COCOM — a body that included all of the NATO member nations, plus Japan and Australia — to present a unified front. By this point, with the Cold War heading into its last series of dangerous crises thanks to Reagan’s bellicosity and the Soviet invasion of Afghanistan, the United States in particular was developing a real paranoia about the Soviet Union’s long-standing habits of industrial espionage. The paranoia was reflected in CIA director William Casey’s testimony to Congress in 1982:

The KGB has developed a large, independent, specialized organization which does nothing but work on getting access to Western science and technology. They have been recruiting about 100 young scientists and engineers a year for the last 15 years. They roam the world looking for technology to pick up. Back in Moscow, there are 400 to 500 assessing what they might need and where they might get it — doing their targeting and then assessing what they get. It’s a very sophisticated and far-flung organization.

By the mid-1980s, restrictions on Western computer exports to the East were quite draconian, a sometimes bewildering maze of regulations to be navigated: 8-bit microcomputers could be exported but 16-bit microcomputers couldn’t be; a single-user accounting package could be exported but not a multi-user version; a monochrome monitor could be exported but not a color monitor.

Even as the barriers between East and West were being piled higher than ever, Western fascination with the Mirror World remained stronger than ever. In August of 1983, an American eye surgeon named Leo D. Bores, organizer of the first joint American/Soviet seminar in medicine in Moscow and a computer hobbyist in his spare time, had an opportunity to spend a week with what was billed as the first ever general-purpose Soviet microcomputer. It was called the “Agat” — just a pretty name, being Russian for the mineral agate — and it was largely a copy — in Bores’s words a bad copy — of the Apple II. His report, appearing belatedly in the November 1984 issue of Byte magazine, proved unexpectedly popular among the magazine’s readership.

The Agat computer

The Agat was, first of all, much, much bigger and heavier than a real Apple II; Bores generously referred to it as “robust.” It was made in a factory more accustomed to making cars and trucks, and, indeed, it looked much as one might imagine a computer built in an automotive plant would look. The Soviets had provided software for displaying text in Cyrillic, albeit with some amount of flicker, using the Apple II’s bitmap-graphics modes. The keyboard also offered Cyrillic input, thus solving, after a fashion anyway, a big problem in adapting Western technology to Soviet needs. But that was about the extent to which the Agat impressed. “The debounce circuitry [on the keyboard] is shaky,” noted Bores, “and occasionally a stray character shows up, especially during rapid data entry. The elevation of the keyboard base (about 3.5 centimeters) and the slightly steeper-than-normal board angle would cause rapid fatigue as well as wrist pain after prolonged use.” Inside the case was a “nightmarish wiring maze.” Rather than being built into a single motherboard, the computer’s components were all mounted on separate breadboards cobbled together by all that cabling, the way Western engineers worked only in the very early prototyping stage of hardware development. The Soviet clone of the MOS 6502 chip found at the heart of the Agat was as clumsily put together as the rest of the machine, spanning across several breadboards; thus this “first Soviet microcomputer” arguably wasn’t really a microcomputer at all by the strict definition of the term. The kicker was the price: about $17,000. As that price would imply, the Agat wasn’t available to private citizens at all, being reserved for use in universities and other centers of higher learning.

With the Cold War still going strong, Byte‘s largely American readership was all too happy to jeer at this example of Soviet backwardness, which certainly did show a computer industry lagging years behind the West. That said, the situation wasn’t quite as bad as Bores’s experience would imply. It’s very likely that the machine he used was a pre-production model of the Agat, and that many of the problems he encountered were ironed out in the final incarnation.

For all the engineering challenges, the most important factor impeding truly personal computing in the Soviet Union was more ideological than technical. As so many of the visionaries who had built the first PCs in the West had so well recognized, these were tools of personal empowerment, of personal freedom, the most exciting manifestation yet of Norbert Wiener’s original vision of cybernetics as a tool for the betterment of the human individual. For an Eastern Bloc still tossing and turning restlessly under the blanket of collectivism, this was anathema. Poland’s propaganda ministry made it clear that they at least feared the existence of microcomputers far more than they did their absence: “The tendency in the mass-proliferation of computers is creating a variety of ideological endangerments. Some programmers, under the inspiration of Western centers of ideological subversion, are creating programs that help to form anti-communistic political consciousness.” In countries like Poland and the Soviet Union, information freely exchanged could be a more potent weapon than any bomb or gun. For this reason, photocopiers had been guarded with the same care as military hardware for decades, and even owning a typewriter required a special permit in many Warsaw Pact countries. These restrictions had led to the long tradition of underground defiance known euphemistically simply as “samizdat,” or self-publishing: the passing of “subversive” ideas from hand to hand as one-off typewritten or hand-written texts. Imagine what a home computer with a word processor and a printer could mean for samizdat. The government of Romania was so terrified by the potential of the computer for spreading freedom that it banned the very word for a time. Harry R. Meyer, an American Soviet watcher with links to the Russian expatriate community, made these observations as to the source of such terror:

I can imagine very few things more destructive of government control of information flow than having a million stations equivalent to our Commodore 64 randomly distributed to private citizens, with perhaps a thousand in activist hands. Even a lowly Commodore 1541 disk drive can duplicate a 160-kilocharacter disk in four or five minutes. The liberating effect of not having to individually enter every character every time information is to be shared should dramatically increase the flow of information.

Information distributed in our society is mainly on paper rather than magnetic media for reasons of cost-effectiveness: the message gets to more people per dollar. The bottleneck of samizdat is not money, but time. If computers were available at any cost, it would be more effective to invest the hours now being spent in repetitive typing into earning cash to get a computer, no matter how long it took.

If I were circulating information the government didn’t like in the Soviet Bloc, I would have little interest in a modem — too easily monitored. But there is a brisk underground trade in audio cassettes of Western music. Can you imagine the headaches (literal and figurative) for security agents if text files were transported by overwriting binary onto one channel in the middle of a stereo cassette of heavy-metal music? One would hope it would be less risk to carry such a cassette than a disk, let alone a compromising manuscript.

If we accept Meyer’s arguments, there’s an ironic follow-on argument to be made: that, in working so hard to keep the latest versions of these instruments of freedom out of the hands of the Soviet Union and its vassal states, the COCOM was actually hurting rather than helping the cause of freedom. As many a would-be autocrat has learned to his dismay in the years since, it’s all but impossible to control the free flow of information in a society with widespread access to personal-computing technology. The new dream of personal computing, of millions of empowered individuals making things and communicating, stood in marked contrast to the Soviet cyberneticists’ old dream of perfect, orderly, top-down control implemented via big mainframe computers. For the hard-line communists, the dream of personal computing sounded more like a nightmare. The Soviet Union faced a stark dilemma: embrace the onrushing computer age despite the loss of control it must imply, or accept that it must continue to fall further and further behind the West. A totalitarian state like the Soviet Union couldn’t survive alongside the free exchange of ideas, while a modern economy couldn’t survive without the free exchange of ideas.

Thankfully for everyone involved, a man now stepped onto the stage who was willing to confront the seemingly insoluble contradictions of Soviet society. On March 11, 1985, Mikhail Gorbachev was named General Secretary of the Communist Party, the eighth and, as it would transpire, the last leader of the Soviet Union. He almost immediately signaled a new official position toward computing, as he did toward so many other things. In one of his first major policy speeches just weeks after assuming power, Gorbachev announced a plan to put personal computers into every classroom in the Soviet Union.

Unlike the General Secretaries who had come before him, Gorbachev recognized that the problems of rampant corruption and poor economic performance which had dogged the Soviet Union throughout its existence were not obstacles external to the top-down collectivist state envisioned by Vladimir Lenin but its inevitable results. “Glasnost,” the introduction of unprecedented levels of personal freedom, and “Perestroika,” the gradual replacement of the planned economy with a more market-oriented version permitting a degree of private ownership, were his responses. These changes would snowball in a way that no one — certainly not Gorbachev himself — had quite anticipated, leading to the effective dissolution of the Warsaw Pact and the end of the Cold War before the 1980s were over. Unnerved by it all though he was, Gorbachev, to his everlasting credit, let it happen, rejecting the calls for a crackdown like those that had ended the Hungarian Revolution of 1956 and the Prague Spring of 1968 in such heartbreak and tragedy.

The Elektronika BK 0010

Very early in Gorbachev’s tenure, well before its full import had even started to become clear, it became at least theoretically possible for the first time for individuals in the Soviet Union to buy a private computer of their own for use in the home. Said opportunity came in the form of the Elektronika BK-0010. Costing about one-fifth as much as the Agat, the BK-0010 was a predictably slapdash product in some areas, such as its horrid membrane keyboard. In other ways, though, it impressed far more than anyone had a right to expect. The BK-0010, the very first Soviet microcomputer designed to be a home computer, was a 16-bit machine, placing it in this respect at least ahead of the typical Western Apple II, Commodore 64, or Sinclair Spectrum of the time. The microprocessor inside it was a largely original creation, borrowing the instruction set from the DEC PDP-11 line of minicomputers but borrowing its actual circuitry from no one. The Soviets’ struggles to stamp out the ever denser circuitry of the latest Western CPUs in their obsolete factories was ironically forcing them to be more innovative, to start designing chips of their own which their factories could manage to produce.

Supplies of the BK-0010 were always chronically short and the waiting lists long, but as early as 1985 a few lucky Soviet households could boast real, usable computers. Those who were less lucky might be able to build a bare-bones computer from schematics published in do-it-yourself technology magazines like Tekhnika Molodezhi, the Soviet equivalent to Popular Electronics. Just as had happened in the United States, Britain, and many other Western countries, a vibrant culture of hobbyist computing spread across the Soviet Union and the other Warsaw Pact nations. In time, as the technology advanced in rhythm with Perestroika, these hobbyists would become the founding spirits of a new Soviet computer industry — a capitalist computer industry. “These are people who have felt useless — useless — all their lives!” said American business pundit Esther Dyson after a junket to a changing Eastern Europe. “Do you know what it is like to feel useless all your life? Computers are turning many of these people into entrepreneurs. They are creating the entrepreneurs these countries need.” As one glance at the flourishing underground economy of the Soviet Union of any era had always been enough to prove, Russians had a natural instinct for capitalism. Now, they were getting the chance to exercise it.

In August of 1988, in a surreal sign of these changing times, a delegation including many senior members of the Soviet Academy of Sciences — the most influential theoretical voice in Soviet computing dating back to the early 1950s — arrived in New York City on a mission that would have been unimaginable just a couple of years before. To a packed room of technology journalists — the Mirror World remained as fascinating as ever — they demonstrated a variety of software which they hoped to sell to the West: an equation solver; a database responsive to natural-language input; a project manager; an economic-modelling package. Byte magazine called the presentation “clever, flashy, and unabashedly commercial,” with “lots of colored windows popping up everywhere” and lots of sound effects. The next few years would bring several ventures which served to prove to any doubters from that initial gathering that the Soviets were capable of programming world-class software if given half a chance. In 1991, for instance, Soviet researchers sold a system of handwriting recognition to Apple for use in the pioneering Apple Newton personal digital assistant. Reflecting the odd blend of greed and idealism that marked the era, a Russian programmer wrote to Byte magazine that “I do hope the world software market will be the only battlefield for American and Soviet programmers and that we’ll become friends during this new battle now that we’ve stopped wasting our intellects on the senseless weapons race.”

As it would transpire, though, the greatest Russian weapon in this new era of happy capitalism wasn’t a database, a project manager, or even a handwriting-recognition system. It was instead a game — a piece of software far simpler than any of those aforementioned things but with perhaps more inscrutable genius than all of them put together. Its unlikely story is next.

(Sources: the academic-journal articles “Soviet Computing and Technology Transfer: An Overview” by S.E. Goodman, “InterNyet: Why the Soviet Union Did Not Build a Nationwide Computer Network” by Slava Gerovitch, “The Soviet Bloc’s Unified System of Computers” by N.C. Davis and S.E. Goodman; the January 1970 and May 1972 issues of Rand Corporation’s Soviet Cybernetics Review; The New York Times of August 28 1966, May 7 1973, and November 27 1979; Scientific American of October 1970; Bloomberg Businessweek of November 4 1991; Byte of August 1980, April 1984, November 1984, July 1985, November 1986, February 1987, October 1988, and April 1989; a video recording the Computer History Museum’s commemoration of the IBM System/360 on April 7 2004. Finally, my huge thanks to Peter Sovietov, who grew up in the Soviet Union of the 1980s and the Russia of the 1990s and has been an invaluable help in sharing his memories and his knowledge and saving me from some embarrassing errors.)

Back in my younger days, when the thought of sleeping for nights on end in campground tents and hostel cots awakened a spirit of adventure instead of a premonition of an aching back, I used to save up my vacation time and undertake a big backpacker-style journey every summer. In 2002, this habit took me to Russia.

I must confess that I found St. Petersburg and Moscow a bit of a disappointment. They just struck me as generic big cities of the sort that I’d seen plenty of in my life. While I’m sure they have their unique qualities, much of what I saw there didn’t look all that distinct from what one could expect to see in any of dozens of major European cities. What I was looking for was the Russia — or, better said, the Soviet Union — of my youth, that semi-mythical Mirror World of fascination and nightmare.

I could feel myself coming closer to my goal as soon as I quit Moscow to board the Trans-Siberian Railroad for the long, long journey to Vladivostok. As everyone who lived in Siberia was all too happy to tell me, I was now experiencing the real Russia. In the city of Ulan-Ude, closed to all outsiders until 1991, I found the existential goal I hadn’t consciously known I’d been seeking. From the central square of Ulan-Ude, surrounded on three sides by government offices still bearing faded hammers and sickles on their facades, glowered a massive bust of Vladimir Lenin. I’d later learn that at a weight of 42 tons the bust was the largest such ever built in the Soviet Union, and that it had been constructed in 1971 as one of the last gasps of the old tradition of Stalinist monumentalism. But the numbers didn’t matter on that scorching-hot summer day when I stood in that square, gazing up in awe. In all my earlier travels, I’d never seen a sight so alien to me. This was it, my personal Ground Zero of the Mirror World, where all the values in which I’d been indoctrinated as a kid growing up deep in the heart of Texas were flipped. Lenin was the greatest hero the world had ever known, the United States the nation of imperialist oppression… it was all so wrong, and because of that it was all so right. I’ve never felt so far from home as I did on that day — and this feeling, of course, was exactly the reason I’d come.

I’m a child of the 1980s, the last decade during which the Soviet Union was an extant power in the world. The fascination which I still felt so keenly in 2002 had been a marked feature of my childhood. Nothing, after all, gives rise to more fascination than telling people that something is forbidden to them, as the Kremlin did by closing off their country from the world. Certainly I wasn’t alone in jumping after any glimpse I could get behind the Iron Curtain.

Thus the bleakly alluring version of Moscow found in Martin Cruz Smith’s otherwise workmanlike crime novel Gorky Park turned it into a bestseller, and then a hit film a couple of years later. (I remember the film well because it was the first R-rated movie my parents ever allowed me to see; I remember being intrigued and a little confused by my first glimpse of bare breasts on film — as if the glimpse behind the Iron Curtain wasn’t attraction enough!) And when David Willis, an American journalist who had lived several years in Moscow, purported to tell his countrymen “how Russians really live” in a book called Klass, it too became a bestseller. Even such a strident American patriot as Tom Clancy could understand the temptation of the Mirror World. In Red Storm Rising, his novel of World War III, straitlaced intelligence officer Robert Toland gets a little too caught up in the classic films of Sergei Eisenstein.

The worst part of the drive home was the traffic to the Hampton Roads tunnel, after which things settled down to the usual superhighway ratrace. All the way home, Toland’s mind kept going over the scenes from Eisenstein’s movie. The one that kept coming back was the most horrible of all, a German knight wearing a crusader’s cross tearing a Pskov infant from his mother’s breast and throwing him — her? — into a fire. Who could see that and not be enraged? No wonder the rabble-rousing song “Arise, you Russian People” had been a genuinely popular favorite for years. Some scenes cried out for bloody revenge, the theme for which was Prokofiev’s fiery call to arms. Soon he found himself humming the song. A real intelligence officer you are … Toland smiled to himself, thinking just like the people you’re supposed to study … defend our fair native land … za nashu zyemlyu chestnuyu!

“Excuse me, sir?” the toll collector asked.

Toland shook his head. Had he been singing aloud? He handed over the seventy-five cents with a sheepish grin. What would this lady think, an American naval officer singing in Russian?

Those involved with computers were likewise drawn to the Mirror World. When Byte magazine ran a modest piece buried hundreds of pages deep in their November 1984 issue on a Soviet personal computer showing the clear “influence” of the Apple II, it became the second most popular article in the issue according to the magazine’s surveys. Unsurprisingly in light of that reception, similar tantalizing glimpses behind the Iron Curtain became a regular part of the magazine from that point forward. According to the best estimates of the experts, the Soviets remained a solid three years behind the United States in their top-end chip-fabrication capabilities, and much further behind than that in their ability to mass-produce dependable computers that could be sold for a reasonable price. If the rudimentary Soviet computers Byte described had come from anywhere else, in other words, no one would have glanced at them twice. Yet the fact that they came from the Mirror World gave them the attraction that clung to all glimpses into that fabled land. For jaded veterans grown bored with an American computer industry that was converging inexorably from the Wild West that had been its early days toward a few standard, well-defined — read, boring — platforms, Soviet computers were the ultimate exotica.

Before the end of the 1980s, an odd little game of falling blocks would ride this tidal wave of Soviet chic to become by some measures the most popular videogame of all time. An aura of inscrutable otherness clung to Tetris, which the game’s various publishers — its publication history is one of the most confusing in the history of videogames — were smart enough to tie in with the sense of otherness that surrounded the entirety of the Soviet Union, the game’s unlikely country of origin, in so many Western minds. Spectrum Holobyte, the most prominent publisher of the game on computers, wrote the name in Cyrillic script on the box front, subtitled it “the Soviet Challenge,” and commissioned background graphics showing iconic — at least to Western eyes — Soviet imagery, from Cosmonauts in space to the “Red Machine” hockey team on the ice. As usual, Nintendo cut more to the chase with their staggeringly successful Game Boy version: “From Russia with Fun!”

Tetris mania was at its peak as the 1990s began. The walls were coming down between West and East, both figuratively and literally, thanks to Mikhail Gorbachev’s impossibly brave choice to let his empire go — peacefully. Western eyes peered eagerly eastward, motivated now not only by innocent if burning curiosity but by the possibilities for tapping those heretofore untapped markets. Having reached this very point here in this blog’s overarching history of interactive entertainment and matters related, let’s hit pause long enough to join those first Western discoverers now in exploring the real story of computing in the Mirror World.

In the very early days of computing, before computer science was a recognized discipline in which you could get a university degree, the most important thinkers in the nascent field tended to be mathematicians. It was, for instance, the British mathematician Alan Turing who laid much of the groundwork for modern computer science in the 1930s, then went on to give many of his theories practical expression as part of the Allied code-breaking effort that did so much to win World War II. And it was the mathematics department of Cambridge University who built the EDSAC in 1949, the first truly programmable computer in the sense that we understand that term today.

The strong interconnection between mathematics and early work with computers should have left the Soviet Union as well-equipped for the dawning age as any nation. Russia had a long, proud tradition of mathematical innovation, dating back through centuries of Czarist rule. The list of major Russian mathematicians included figures like Nikolai Lobachevsky, the pioneer of non-Euclidean geometry, and Sofia Kovalevskaya, who developed equations for the rotation of a solid body around a fixed axis. Even Joseph Stalin’s brutal purges of the 1930s, which strove to expunge anyone with the intellectual capacity to articulate a challenge to his rule, failed to kill the Russian mathematical tradition. On the contrary, Leonid Kantorovich in 1939 discovered the technique of linear programming ten years before American mathematicians would do the same, while Andrey Kolmogorov did much fundamental work in probability theory and neural-network modeling over a long career that spanned from the 1920s through the 1980s. Indeed, in the decades following Stalin’s death, Soviet mathematicians in general would continue to solve fundamental problems of theory. And Soviet chess players — the linkage between mathematics and chess is almost as pronounced in history as that between mathematics and computers — would remain the best in the world, at least if the results of international competitions were any guide.

But, ironically in light of all this, it would be an electrical engineer named Sergei Alexeevich Lebedev rather than a mathematician who would pioneer Soviet computing. Lebedev was 46 years old in 1948 when he was transferred from his cushy position at the Lenin State Electrical Institute in Moscow to the relative backwater of Kiev, where he was to take over as head of the Ukraine Academy’s Electrotechnical Institute. There, free from the scrutiny of Moscow bureaucrats who neither understood nor wanted to understand the importance of the latest news of computing coming out of Britain and the United States, Lebedev put together a small team to build a Small Computing Machine; in Russian its acronym was MESM. Unlike the team of scientists and engineers who detonated the Soviet Union’s first atomic bomb in 1949, Lebedev developed the MESM without the assistance of espionage; he had access to the published papers of figures like Alan Turing and the exiled Hungarian mathematician John von Neumann, but no access to schematics or inside information about the machines on which they were working.

Lebedev had to build the MESM on a shoestring. Just acquiring the vacuum tubes and magnetic drums he needed in a backwater city of a war-devastated country was a major feat in itself, one that called for the skills of a junk trader as much as it did those of an electrical engineer. Seymour Goodman, one of the more notable historians of Soviet computing, states that “perhaps the most incredible aspect of the MESM was that it was successfully built at all. No electronic computer was ever built under more difficult conditions.” When it powered up for the first time in 1951, the MESM was not only the first stored-program computer in the Soviet Union but the first anywhere in continental Europe, trailing Britain by just two years and the United States by just one — a remarkable achievement by any standard.

Having already shown quite a diverse skill set in getting the MESM made at all, Lebedev proved still more flexible after it was up and running. He became the best advocate for computing inside the Soviet Union, a sort of titan of industry in a country that officially had no room for such figures. Goodman credits him with playing the role that a CEO would have played in the West. He even managed to get a script written for a documentary film to “advertise” his computer’s capabilities throughout the Soviet bureaucracy. In the end, the film never got made, but then it really wasn’t needed. The Soviet space and nuclear-weapons programs, not to mention the conventional military, all had huge need of the fast calculations the MESM could provide. At the time, the nuclear-weapons program was using what they referred to as calculator “brigades,” consisting of 100 or more mostly young girls, who worked eight-hour shifts with mechanical devices to crank out solutions to hugely complicated equations. Already by 1950, an internal report had revealed that the chief obstacle facing Soviet nuclear scientists wasn’t the theoretical physics involved but rather an inability to do the math necessary to bring theory to life fast enough.

Within months of his machine going online, Lebedev was called back to Moscow to become the leader of the Institute for Precision Mechanics and Computing Technology — or ITMVT in the Russian acronym — of the Soviet Academy of Sciences. There Lebedev proceeded to develop a series of machines known as the BESM line, which, unlike the one-off MESM, were suitable for — relatively speaking — production in quantity.

But Lebedev soon had rivals. Contrary to the image the Kremlin liked to project of a unified front — of comrades in communism all moving harmoniously toward the same set of goals — the planned economy of the Soviet Union was riddled with as much in-fighting as any other large bureaucracy. “Despite its totalitarian character,” notes historian Nikolai Krementsov, “the Soviet state had a very complex internal structure, and the numerous agents and agencies involved in the state science-policy apparatus pursued their own, often conflicting policies.” Thus very shortly after the MESM became operational, the second computer to be built in the Soviet Union (and continental Europe as well), a machine called the M-1 which had been designed by one Isaak Semyenovich Bruk, went online. If Lebedev’s achievement in building the MESM was remarkable, Bruk’s achievement in building the M-1, again without access to foreign espionage — or for that matter the jealously guarded secrets of Lebedev’s rival team — was equally so. But Bruk lacked Lebedev’s political skills, and thus his machine proved a singular achievement rather than the basis for a line of computers.

A much more dangerous rival  was a computer called Strela, or “Arrow,” the brainchild of one Yuri Yakovlevich Bazilevskii in the Special Design Bureau 245 — abbreviated SKB-245 in Russian — of the Ministry of Machine and Instrument Construction in Moscow. The BESM and Strela projects, funded by vying factions within the Politburo, spent several years in competition with one another, each project straining to monopolize scarce components, both for its own use and, just as importantly, to keep them out of the hands of its rival. It was a high-stakes war that was fought in deadly earnest, and its fallout could be huge. When, for instance, the Strela people managed to buy up the country’s entire supply of cathode-ray tubes for use as memory, the BESM people were forced to use less efficient and reliable mercury delay lines instead. As anecdotes like this attest, Bazilevskii was every bit Lebedev’s equal at the cutthroat game of bureaucratic politicking, even managing to secure from his backers the coveted title of Hero of Socialist Labor a couple of years before Lebedev.

The Strela computer. Although it’s hard to see it here, it was described by its visitors as a “beautiful machine in a beautiful hall,” with hundreds of lights blinking away in impressive fashion. Many bureaucrats likely chose to support the Strela simply because it looked so much like the ideal of high technology in the popular imagination of the 1950s.

During its first official trial in the spring of 1954, the Strela solved in ten hours a series of equations that would have taken a single human calculator about 100,000 days. And the Strela was designed to be a truly mass-produced computer, to be cranked out in the thousands in identical form from factories. But, as so often happened in the Soviet Union, the reality behind the statistics which Pravda trumpeted so uncritically was somewhat less flattering. The Strela “worked very badly” according to one internal report; according to another it “very often failed and did not work properly.” Pushed by scientists and engineers who needed a reliable computer in order to get things done, the government decided in the end to go ahead with the BESM instead of the Strela. Ironically, only seven examples of the first Soviet computer designed for true mass-production were ever actually produced. Sergei Lebedev was now unchallenged as the preeminent voice in Soviet computing, a distinction he would enjoy until his death in 1974.

The first BESM computer. It didn’t look as nice as the Strela, but it would prove far more capable and reliable.

Like so much other Soviet technology, Soviet computers were developed in secrecy, far from the prying eyes of the West. In December of 1955, a handful of American executives and a few journalists on a junket to the Soviet Union became the first to see a Soviet computer in person. A report of the visit appeared in the New York Times of December 11, 1955. It helpfully describes an early BESM computer as an “electronic brain” — the word “computer” was still very new in the popular lexicon — and pronounces it equal to the best American models of same. In truth, the American delegation had fallen for a bit of a dog-and-pony show. Soviet computers were already lagging well behind the American models that were now being churned out in quantities Lebedev could only dream of by companies like IBM.

Sergei Lebedev’s ITMVT. (Sorry for the atrocious quality of these images. Clear pictures of the Mirror World of the 1950s are hard to come by.)

In May of 1959, during one of West and East’s periodic periods of rapprochement, a delegation of seven American computer experts from business and government was invited to spend two weeks visiting most of the important hubs of computing research in the Soviet Union. They were met at the airport in Moscow by Lebedev himself; the Soviets were every bit as curious about the work of their American guests as said Americans were about theirs. The two most important research centers of all, the American delegation learned, were Lebedev’s ITMVT and the newer Moscow Computing Center of the Soviet Academy of Sciences, which was coming to play a role in software similar to that which the ITMVT played in hardware. The report prepared by the delegation is fascinating for the generalized glimpses it provides into the Soviet Mirror World of the 1950s as much as it is for the technical details it includes. Here, for instance, is its description of the ITMVT’s physical home:

The building itself is reminiscent more of an academic building than an industrial building. It is equipped with the usual offices and laboratory facilities as well as a large lecture hall. Within an office the decor tends to be ornate; the entrance door is frequently padded on both sides with what appeared to be leather, and heavy drapery is usually hung across the doorway and at the windows. The ceiling height was somewhat higher than that of contemporary American construction, but we felt in general that working conditions in the offices and in the laboratories were good. There appeared to be an adequate amount of room and the workers were comfortably supplied with material and equipment. The building was constructed in 1951. Many things testified to the steady and heavy usage it has received. In Russian tradition, the floor is parqueted and of unfinished oak. As in nearly every building, there are two sets of windows for weather protection.

The Moscow Computing Center

And here’s how a Soviet programmer had to work:

Programmers from the outside who come to the [Moscow] Computing Center with a problem apply to the scientific secretary of the Computing Center. He assigns someone from the Computing Center to provide any assistance needed by the outside programmer. In general an operator is provided for each machine, and only programmers with specific permission can operate the machine personally. Normally a programmer can expect only one code check pass per day at a machine; with a very high priority he might get two or three passes.

A programmer is required to submit his manuscript in ink. Examples of manuscripts which we saw indicated that often a manuscript is written in pencil until it is thought to be correct, and then redone in ink. The manuscript is then key-punched twice, and the two decks compared, before being sent to the machine. The output cards are handled on an off-line printer.

Other sections describe the Soviet higher-education system (“Every student is required to take 11 terms of ideological subjects such as Marxism-Leninism, dialectical materialism, history of the Communist Party, political economy, and economics.”); the roles of the various Academies of Sciences (“The All Union Academy of Sciences of the USSR and the 15 Republican Academies of Sciences play a dominant role in the scientific life of the Soviet Union.”); the economics of daily life (“In evaluating typical Russian salaries it must be remembered that the highest income tax in the Soviet Union is 13 percent and that all other taxes are indirect.”); the resources being poured into the new scientific and industrial center of Novosibirsk (“It is a general belief in Russia that the future of the Soviet Union is closely allied with the development of the immense and largely unexplored natural resources of Siberia.”).

But of course there are also plenty of pages devoted to technical discussion. What’s most surprising about these is the lack of the hysteria that had become so typical of Western reports of Soviet technology in the wake of the Sputnik satellite of 1957 and the beginning of the Space Race which it heralded. It was left to a journalist from the New York Times to ask the delegation upon their return the money question: who was really ahead in the field of computers? Willis Ware, a member of the delegation from the Rand Corporation and the primary architect of the final report, replied that the Soviet Union had “a wealth of theoretical knowledge in the field,” but “we didn’t see any hardware that we don’t have here.” Americans had little cause to worry; whatever their capabilities in the fields of aerospace engineering and nuclear-weapons delivery, it was more than clear that the Soviets weren’t likely to rival even IBM alone, much less the American computer industry as a whole, anytime soon. With that worry dispensed with, the American delegation had felt free just to talk shop with their Soviet counterparts in what would prove the greatest meeting of Eastern and Western computing minds prior to the Gorbachev era. The Soviets responded in kind; the visit proved remarkably open and friendly.

One interesting fact gleaned by the Americans during their visit was that, in addition to all the differences born of geography and economy, the research into computers conducted in the East and the West had also heretofore had markedly different theoretical scopes. For all that so much early Western research had been funded by the military for such plebeian tasks as code-breaking and the calculation of artillery trajectories, and for all that so much of that research had been conducted by mathematicians, the potential of computers to change the world had always been understood by the West’s foremost visionaries as encompassing far more than a faster way to do complex calculations. Alan Turing, for example, had first proposed his famous Turing Test of artificial intelligence all the way back in 1950.

But in the Soviet Union, where the utilitarian philosophy of dialectical materialism was the order of the day, such humanistic lines of research were, to say the least, not encouraged. Those involved with Soviet computing had to be, as they themselves would later put it, “cautious” about the work they did and the way they described that work to their superiors. The official view of computers in the Soviet Union during the early and mid-1950s hewed to the most literal definition of the word: they were electronic replacements for those brigades of human calculators cranking out solutions to equations all day long. Computers were, in other words, merely a labor-saving device, not a revolution in the offing; being a state founded on the all-encompassing ideology of communist revolution, the Soviet Union had no use for other, ancillary revolutions. Even when Soviet researchers were allowed to stray outside the realm of pure mathematics, their work was always expected to deliver concrete results that served very practical goals in fairly short order. For example, considerable effort was put into a program for automatically translating texts between languages, thereby to better bind together the diverse peoples of the sprawling Soviet empire and its various vassal states. (Although the translation program was given a prominent place in that first 1955 New York Times report about the Soviets’ “electronic brain,” one has to suspect that, given how difficult a task automated translation is even with modern computers, it never amounted to much more than a showpiece for use under carefully controlled conditions.)

And yet even by the time the American delegation arrived in 1959 all of that was beginning to change, thanks to one of the odder ideological alliances in the history of the twentieth century. In a new spirit of relative openness that was being fostered by Khrushchev, the Soviet intelligentsia was becoming more and more enamored with the ideas of an American named Norbert Wiener, yet another of those wide-ranging mathematicians who were doing so much to shape the future. In 1948, Wiener had described a discipline he called “cybernetics” in a book of the same name. The book bore the less-than-enticing subtitle Control and Communication in the Animal and the Machine, making it sound rather like an engineering text. But if it was engineering Wiener was practicing, it was social engineering, as became more clear in 1950, when he repackaged his ideas into a more accessible book with the title The Human Use of Human Beings.

Coming some 35 years before William Gibson and his coining of the term “cyberspace,” Norbert Wiener marks the true origin point of our modern mania for all things “cyber.” That said, his ideas haven’t been in fashion for many years, a fact which might lead us to dismiss them from our post-millennial perch as just another musty artifact of the twentieth century and move on. In actuality, though, Wiener is well worth revisiting, and with an eye to more than dubious linguistic trends. Cybernetics as a philosophy may be out of fashion, but cybernetics as a reality is with us a little more every day. And, most pertinently for our purposes today, we need to understand a bit of what Wiener was on about if we hope to understand what drove much of Soviet computing for much of its existence.

“Cybernetics” is one of those terms which can seem to have as many definitions as definers. It’s perhaps best described as the use of machines not just to perform labor but to direct labor. Wiener makes much of the increasing numbers of machines even in his time which incorporated a feedback loop — machines, in other words, that were capable of accepting input from the world around them and responding to that input in an autonomous way. An example of such a feedback loop can be something as simple as an automatic door which opens when it senses people ready to step through it, or as complex as the central computer in charge of all of the functions of an automated factory.

At first blush, the idea of giving computers autonomous control over the levers of power inevitably conjures up all sorts of dystopian visions. Yet Wiener himself was anything but a fan of totalitarian or collectivist governments. Invoking in The Human Use of Human Beings the popular metaphor of the collectivist society as an ant colony, he goes on to explore the many ways in which humans and ants are in fact — ideally, at any rate — dissimilar, thus seemingly exploding the “from each according to his ability, to each according to his need” founding principle of communism.

In the ant community, each worker performs its proper functions. There may be a separate caste of soldiers. Certain highly specialized individuals perform the functions of king and queen. If man were to adopt this community as a pattern, he would live in a fascist state, in which ideally each individual is conditioned from birth for his proper occupation: in which rulers are perpetually rulers, soldiers perpetually soldiers, the peasant is never more than a peasant, and the worker is doomed to be a worker.

This aspiration of the fascist for a human state based on the model of the ant results from a profound misapprehension both of the nature of the ant and of the nature of man. I wish to point out that the very physical development of the insect conditions it to be an essentially stupid and unlearning individual, cast in a mold which cannot be modified to any great extent. I also wish to show how these physiological conditions make it into a cheap mass-produced article, of no more individual value than a paper pie plate to be thrown away after it is used. On the other hand, I wish to show that the human individual, capable of vast learning and study, which may occupy almost half his life, is physically equipped, as the ant is not, for this capacity. Variety and possibility are inherent in the human sensorium — and are indeed key to man’s most noble flights — because variety and possibility belong to the very structure of the human organism.

While it is possible to throw away this enormous advantage that we have over the ants, and to organize the fascist ant-state with human material, I certainly believe that this is a degradation of man’s very nature, and economically a waste of the great human values which man possesses.

I am afraid that I am convinced that a community of human beings is a far more useful thing than a community of ants, and that if the human being is condemned and restricted to perform the same functions over and over again, he will not even be a good ant, not to mention a good human being. Those who would organize us according to personal individual functions and permanent individual restrictions condemn the human race to move at much less than half-steam. They throw away nearly all our human possibilities and, by limiting the modes in which we may adapt ourselves to future contingencies, they reduce our chances for a reasonably long existence on this earth.

Wiener’s vision departs markedly from the notion, popular already in science fiction by the time he wrote those words, of computers as evil overlords. In Wiener’s cybernetics, computers will not enslave people but give them freedom; the computers’ “slaves” will themselves be machines. Together computers and the machines they control will take care of all the boring stuff, as it were, allowing people to devote themselves to higher purposes. Wiener welcomes the “automatic age” he sees on the horizon, even as he is far from unaware of the disruptions the period of transition will bring.

What can we expect of its economic and social consequences? In the first place, we can expect an abrupt and final cessation of the demand for the type of factory labor performing purely repetitive tasks. In the long run, the deadly uninteresting nature of the repetitive task may make this a good thing and the source of leisure necessary for man’s full cultural development.

Be that as it may, the intermediate period of the introduction of the new means will lead to an immediate transitional period of disastrous confusion.

In terms of cybernetics, we’re still in this transitional period today, with huge numbers of workers accustomed to “purely repetitive tasks” cast adrift in this dawning automatic age; this explains much about recent political developments over much of the world. But of course our main interest right now isn’t contemporary politics, but rather how a fellow who so explicitly condemned the collectivist state came to be regarded as something of a minor prophet by the Soviet bureaucracy.

Wiener’s eventual acceptance in the Soviet Union is made all the more surprising by the Communist Party’s first reaction to cybernetics. In 1954, a year after Stalin’s death, the Party’s official Brief Philosophical Dictionary still called cybernetics “a reactionary pseudo-science originating in the USA after World War II and spreading widely in other capitalistic countries as well.” It was “in essence aimed against materialistic dialectics” and “against the scientific Marxist understanding of the laws of societal life.” Seemingly plucking words at random from a grab bag of adjectives, the dictionary concluded that “this mechanistic, metaphysical pseudo-science coexists very well with idealism in philosophy, psychology, and sociology” — the word “idealism” being a kiss of death under Soviet dogma.

In 1960, six years after the Soviets condemned cybernetics as an “attempt to transform toilers into mere appendices of the machine, into a tool of production and war,” Nobert Wiener lectures the Leningrad Mathematical Society. A colleague who visited the Soviet Union at the same time said that Wiener was “wined and dined everywhere, even in the privacy of the homes of the Russian scientists.” He died four years later, just as the influence of cybernetics was reaching a peak in the Soviet Union.

Still, when stripped of its more idealistic, humanistic attributes, there was much about cybernetics which held immense natural appeal for Soviet bureaucrats. Throughout its existence, the Soviet Union’s economy had been guided, albeit imperfectly at best, by an endless number of “five-year plans” that attempted to control its every detail. Given this obsession with economic command and control and the dispiriting results it had so far produced, the prospect of information-management systems — namely, computers — capable of aiding decision-making, or perhaps even in time of making the decisions, was a difficult enticement to resist; never mind how deeply antithetical the idea of computerized overlords making the decisions for human laborers was to Norbert Wiener’s original conception of cybernetics. Thus cybernetics went from being a banned bourgeois philosophy during the final years of Stalin’s reign to being a favorite buzzword during the middle years of Khrushchev’s. In December of 1957, the Soviet Academy of Sciences declared their new official position to be that “the use of computers for statistics and planning must have an absolutely exceptional significance in terms of its efficiency. In most cases, such use would make it possible to increase the speed of decision-making by hundreds of times and avoid errors that are currently produced by the unwieldy bureaucratic apparatus involved in these activities.”

In October of 1961, the new Cybernetics Council of the same body published an official guide called Cybernetics in the Service of Communism — essentially Norbert Wiener with the idealism and humanism filed off. Khrushchev may have introduced a modicum of cultural freedom to the Soviet Union, but at heart he was still a staunch collectivist, as he made clear:

In our time, what is needed is clarity, ideal coordination, and organization of all links in the social system both in material production and in spiritual life.

Maybe you think there will be absolute freedom under communism? Those who think so don’t understand what communism is. Communism is an orderly, organized society. In that society, production will be organized on the basis of automation, cybernetics, and assembly lines. If a single screw is not working properly, the entire mechanism will grind to a halt.

Soviet ambitions for cybernetics were huge, and in different circumstances might have led to a Soviet ARPANET going online years before the American version. It was envisioned that each factory and other center of production in the country would be controlled by its own computer, and that each of these computers would in turn be linked together into “complexes” reporting to other computers, all of which would send their data yet further up the chain, culminating in a single “unified automated management system” directing the entire economy. The system would encompass tens of thousands of computers, spanning the width and breadth of the largest country in the world, “from the Pacific to the Carpathian foothills,” as academician Sergei Sobolev put it. Some more wide-eyed prognosticators said that in time the computerized cybernetic society might allow the government to eliminate money from the economy entirely, long a cherished dream of communism. “The creation of an automated management system,” wrote proponent Anatolii Kitov, “would mean a revolutionary leap in the development of our country and would ensure a complete victory of socialism over capitalism.” With the Soviet Union’s industrial output declining every year between 1959 and 1964 while the equivalent Western figures skyrocketed, socialism needed all the help it could get.

In May of 1962, in an experiment trumpeted as the first concrete step toward socialism’s glorious cybernetic future, a computer located in Kiev poured steel in a factory located hundreds of kilometers away in Dniprodzerzhynsk (known today as Kamianske). A newspaper reporter was inspired to wax poetic:

In ancient Greece the man who steered ships was called Kybernetes. This steersman, whose name is given to one of the boldest sciences of the present — cybernetics — lives on in our own time. He steers the spaceships and governs the atomic installations, he takes part in working out the most complicated projects, he helps to heal humans and to decipher the writings of ancient peoples. As of today he has become an experienced metallurgist.

Some Soviet cybernetic thinking is even more astonishing than their plans for binding the country in a web of telecommunications long before “telecommunications” was a word in popular use. Driverless cars and locomotives were seriously discussed, and experiments with the latter were conducted in the Moscow subway system. (“Experiments on the ‘auto-pilot’ are being concluded. This device, provided with a program for guiding a train, automatically decreases and increases speed at corresponding points along its route, continually selecting the most advantageous speed, and stops the train at the required points.”) Serious attention was given to a question that still preoccupies futurists today: that of the role of human beings in a future of widespread artificially intelligent computers. The mathematician Kolmogorov wrote frankly that such computers could and inevitably would “surpass man in his development” in the course of time, and even described a tipping point that we still regard as seminal today: the point when artificial intelligence begins to “breed,” to create its own progeny without the aid of humans. At least some within the Soviet bureaucracy seemed to welcome humanity’s new masters; proposals were batted around to someday replace human teachers and doctors with computers. Sergei Sobolev wrote that “in my view the cybernetic machines are people of the future. These people will probably be much more accomplished than we, the present people.” Soviet thinking had come a long way indeed from the old conception of computers as nothing more than giant calculators.

But the Soviet Union was stuck in a Catch-22 situation: the cybernetic command-and-control network its economy supposedly needed in order to spring to life was made impossible to build by said economy’s current moribund state. Some skeptical planners drew pointed comparisons to the history of another sprawling land: Egypt. While the Pharaohs of ancient Egypt had managed to build the Pyramids, the cybernetics skeptics noted, legend held that they’d neglected everything else so much in the process that a once-fertile land had become a desert. Did it really make sense to be thinking already about building a computer network to span the nation when 40 percent of villages didn’t yet boast a single telephone within their borders? By the same token, perhaps the government should strive for the more tangible goal of placing a human doctor within reach of every citizen before thinking about replacing all the extant human doctors with some sort of robot.

A computer factory in Kiev, circa 1970. Note that all of the assembly work is still apparently done by hand.

The skeptics probably needn’t have worried overmuch about their colleagues’ grandiose dreams. With its computer industry in the shape it was, it was doubtful whether the Soviet Union had any hope of building its cybernetic Pyramids even with all the government will in the world.

In November of 1964, another American delegation was allowed a glimpse into the state of Soviet computing, although the Cuban Missile Crisis and other recent conflicts meant that their visit was much shorter and more restricted than the one of five and a half years earlier. Regardless, the Americans weren’t terribly impressed by the factory they were shown. It was producing computers at the rate of about seven or eight per month, and the visitors estimated its products to be roughly on par with an IBM 704 — a model that IBM had retired four years before. It was going to be damnably hard to realize the Soviet cybernetic dream with this trickle of obsolete machines; estimates were that about 1000 computers were currently operational in the Soviet Union, as compared to 30,000 in the United States. The Soviets were still struggling to complete the changeover from first-generation computer hardware, characterized by its reliance on vacuum tubes, to the transistor-based second generation. The Americans had accomplished this changeover years before; indeed, they were well on their way to an integrated-circuit-based third generation.  Looking at a Soviet transistor, the delegation said it was roughly equivalent to an American version of same from 1957.

But when the same group visited the academics, they were much more impressed, noting that the Soviets “were doing quite a lot of very good and forward-thinking work.” Thus was encapsulated what would remain the curse of Soviet computer science: plenty of ideas, plenty of abstract know-how, and a dearth of actual hardware to try it all out on. The reports of the Soviet researchers ooze frustration with their lot in life. Their computers break down “each and every day,” reads one, “and information on a tape lasts without any losses no longer than one month.”

Their American visitors were left to wonder just why it was that the Soviet Union struggled so mightily to build a decent computing infrastructure. Clearly the Soviets weren’t complete technological dunces; this was after all the country that had detonated an atomic bomb years before anyone had dreamed it could, that had shocked the world by putting the first satellite and then the first man into space, that was even now giving the United States a run for its money to put a man on the moon.

The best way to address the Americans’ confusion might be to note that exploding atomic bombs and launching things into space encompassed a series of individual efforts responsive to brilliant individual minds, while the mass-production of the standardized computers that would be required to realize the cybernetics dream required a sort of infrastructure-building at which the Soviet system was notoriously poor. The world’s foremost proponent of collectivism was, ironically, not all that good at even the most fundamental long-term collectivist projects. The unstable Soviet power grid was only one example; the builders of many Soviet computer installations had to begin by building their own power plant right outside the computer lab just to get a dependable electrical supply.

The Soviet Union was a weird mixture of backwardness and forwardness in terms of technology, and the endless five-year plans only exacerbated its issues by emphasizing arbitrary quotas rather than results that mattered in the real world. Stories abounded of factories that produced lamp shades in only one color because that was the easiest way to make their quota, or that churned out uselessly long, fat nails because the quota was given in kilograms rather than in numbers of individual pieces. The Soviet computer industry was exposed to all these underlying economic issues. It was hard to make computers to rival those of the West when the most basic electrical components that went into them had failure rates dozens of times higher than their Western equivalents. Whether a planned economy run by computers could have fixed these problems is doubtful in the extreme, but at any rate the Soviet cyberneticists would never get a chance to try. It was the old chicken-or-the-egg conundrum. They thought they needed lots of good computers to build a better economy — but they knew they needed a better economy to build lots of good computers.

As the 1960s became the 1970s, these pressures would lead to a new approach to computer production in the Soviet Union. If they couldn’t beat the West’s computers with their homegrown designs, the Soviets decided, then they would just have to  clone them.

(Sources: the academic-journal articles “Soviet Computing and Technology Transfer: An Overview” by S.E. Goodman, “MESM and the Beginning of the Computer Era in the Soviet Union” by Anne Fitzpatrick, Tatiana Kazakova, and Simon Berkovich, “S.A. Lebedev and the Birth of Soviet Computing” by G.D. Crowe and S.E. Goodman, “The Origin of Digital Computing in Europe” by S.E. Goodman, “Strela-1, The First Soviet Computer: Political Success and Technological Failure” by Hiroshi Ichikawa, and “InterNyet: Why the Soviet Union Did Not Build a Nationwide Computer Network” by Slava Gerovitch; studies from the Rand Corporation entitled “Soviet Cybernetics Technology I: Soviet Cybernetics, 1959-1962” and “Soviet Computer Technology — 1959”; the January 1970 issue of Rand Corporation’s Soviet Cybernetics Review; the books Stalinist Science by Nikolai Krementsov, The Human Use of Human Beings by Norbert Wiener, Red Storm Rising by Tom Clancy, and From Newspeak to Cyberspeak: A History of Soviet Cybernetics by Slava Gerovitch; The New York Times of December 11 1955, December 2 1959, and August 28 1966; Scientific American of October 1970; Byte of November 1984, February 1985, and October 1987.)