The untold story of the computer genius spy who fascinated the Soviets - then mysteriously wound up dead

The Telegraph
By Iain Dey

How was the life of brilliant US scientist Dudley Buck cut short at 32 after a visit by Soviet counterparts? In his new book, The Cryotron Files, Iain Dey reports...

Rainclouds were hanging over Idlewild International Airport as the KLM flight from Amsterdam touched down on the runway. Sergey Lebedev peered out of the window, unimpressed.

They had told him in Moscow that New York was at its best in April – bright and sunny, yet without the oppressive heat and humidity of summer. He had left his raincoat at home and advised the six Soviet computer experts joining him for the trip not to bother bringing theirs either.

The propellers of the Lockheed Super Constellation were still winding down as the group got to the top of the plane’s steps. One by one they looked up at the gathering storm and realised their wardrobe error. Above the noise of the engines, Lebedev could hear the grumbling begin. It was Sunday, April 19 1959. They had left Moscow two days earlier, and all were in need of sleep.

Each man carried a black leather briefcase. Some contained drawings and notes about the biggest and best computers in the Soviet Union – information they planned to present to the Americans. Others were carrying vodka and black caviar to treat their hosts during what was scheduled to be a two-week tour of the United States.

They had come for a rapprochement. The US government had agreed to let the Russians see inside America’s most secret computer labs; the Kremlin would offer the same courtesy in exchange.

Lebedev, at age 56 the Soviet Union’s top computer expert, had been tasked with leading the delegation himself. During the Second World War he had built a system to stabilise the sights of tank cannons. He then created the first computer in the Eastern bloc with a small group of researchers at the Ukrainian Academy of Sciences, which in turn led to him being hand-picked by Joseph Stalin to lead the USSR’s computer effort.

He had retained the role under the new premier, Nikita Khrushchev, and was finally starting to make progress. Although the Soviet Union had caught up with Americans on the nuclear bomb, and had beaten the US into space with the launch of Sputnik some 15 months earlier, computer technology was one area where the Americans had a sizeable advantage.

Stalin was an obstacle to the development of Soviet computer technology. He had objected to the development of any machine that would replicate the human brain or replace a man on a factory production line; he saw it as a capitalist evil. That had forced Lebedev and his contemporaries to develop computers with very strictly defined military missions: for translation, weather forecasting and to calculate the firing range of missiles.

America, on the other hand, had burned billions of dollars on a sprawling mass of computer projects with undefined or moving objectives. Private companies were competing with universities and government departments for lucrative defence contracts to build computers for the army, the air force, the navy, or the newly created commercial honeypot that was Nasa, the American space agency. It was a creative hotbed that had spawned a booming industry, one that was inventing ever more advanced technologies at breakneck speed.

Military might

Lebedev had built an impressive machine in his lab in Moscow, but had not worked out how to mass-produce the device effectively. The Americans, meanwhile, were already rolling out reliable computers by the hundreds. American businesses were installing giant machines sold by the likes of IBM and RCA, which could be used to run their payrolls or settle their taxes. Programmes were under way to computerise air traffic control and US census data.

Both superpowers knew computer technology had the power to change the dynamics of the Cold War. There were clear economic benefits to be gained from the digitisation of the American economy.

Yet there were also more direct military uses for computing power. Both sides were developing nuclear missiles at great pace, and computers were needed to guide those missiles and to identify and shoot down any incoming enemy threats. The American science community was bubbling with stories about one young scientist in particular.

Dudley Buck at MIT had been part of the team that built the first ever random-access memory (Ram) – used in the Project Whirlwind missile defence computer.

He had created an early version of the flash drive, as well as a primitive version of the light gun, such as the one used on the original Eighties Nintendo games console.

Although still just 32, he had won international fame for developing an ultra-fast computer with no moving parts that he repeatedly claimed would “fit in a man’s shirt pocket”. Given that the most advanced computers at that time occupied whole floors of office buildings, it was an attention-grabbing concept. Although the term had not yet been coined, he had invented a prototype microchip – which he named the Cryotron.

According to an article Lebedev had seen in Life magazine two years earlier, Buck’s tiny computer chip would be used as the guidance system for America’s new intercontinental ballistic missile.

At the time the article was published, Buck’s prototype device was a long way from being capable of deployment with a nuclear warhead. In the intervening two years, however, a number of large research projects under the auspices of the US government had been created to drive forward Buck’s Cryotron technology. The US State Department had given Lebedev and his team permission to see inside Buck’s lab. Just three days after Lebedev and his team of scientists touched down in New York, they were scheduled to meet Buck – and to see his invention for themselves.

Buck's double life

To his students, Buck was a gifted young professor who sang along to show tunes in his lab and loved a practical joke. When MIT received its first ever sample of superglue, Buck had used it to stick the janitor’s fingers together. Although a teetotaller himself, with a wife and three young children at home, he had turned a blind eye to an illicit gin still at the back of his lab. He also tolerated the snakes, frogs and scarecrows that turned up on his work benches as part of pranks played by his students.

While his students knew of his inventions, they knew nothing of his double life. To his students, Buck was a gifted young professor who sang along to show tunes in his lab and loved a practical joke. 

As well as an MIT scientist, Buck was a government agent. For the previous nine years he had been working part-time for the National Security Agency and its predecessor organisations, playing roles large and small in classified defence projects – such as the Corona spy satellite programme, early attempts at artificial intelligence and countless schemes to build bigger and better computers for various branches of the military. Many of today’s “big data” computer systems rely on a memory that Buck created to solve one of these problems.

Buck had worked as a codebreaker in Washington, at Csaw – America’s equivalent to Bletchley Park. Diary entries show that he was familiar with many of the Manhattan Project scientists. He had even spent time seconded to one of the most infamous intelligence arms of the CIA, which took him behind enemy lines in Eastern Europe – where he appears to have been involved in attempts to persuade German computer scientist Konrad Zuse to defect to the US.

Throughout his time at MIT, Buck moonlighted as one of the NSA’s top troubleshooters – coming up with solutions to seemingly intractable problems. His Cryotron chip was now the centre of his military life, as well as scientific. Since the USSR had launched its Sputnik satellite 18 months earlier, building better computers had become an obsession of the White House and the Pentagon.

To build a better computer required finding a way to create a device that could switch from an “on” position to an “off” position extremely quickly – from “1” to “0” in terms of the language of binary code upon which all computer programs depend. 

While the earliest computers had used mechanical switches, scientists across the world were now racing to find better, quicker and more efficient electronic switches. For it was only once the switches got quicker that computers would be able to start fulfilling their potential by performing ever more complex tasks.

Buck’s design relied on superconductors – chemical elements that conduct electricity at ultralow temperatures below minus 148F. His experiments had to be suspended inside vats of liquid helium.

He had started by experimenting with cheap superconductors such as lead. Increasingly, however, he was turning to an assortment of rare earth metals that no one in the lab had ever seen before. His favourites were tantalum and niobium. One metal could be used to make the other flip between “1” and “0”.

The Cryotron had started as two small wires wound round one another by hand. As they had perfected their technique, however, Buck and his lab partner Ken Shoulders had developed a much more advanced technique. Using an electron gun, they would lay thin lines of the metals on to a plate – creating, in effect, one of the first integrated circuits – or microchips.

All over the world, scientists were competing to build the first microchip. Many avenues were being pursued, including the semiconducting silicon chip that eventually won the battle and drives most computers today. Yet, at the time, Buck was considered to have the scientific lead with his concept of the “superconducting” microchip. In spring 1959, it was still believed that a silicon chip would melt before it could switch at sufficient speed from one to zero.

Buck had not quite perfected the Cryotron. It still had flaws and was not living up to its potential. None the less, a steady stream of newspaper reporters had trickled through his spartan little office. A scriptwriter had come to interview him about turning the story of his invention into a prime-time drama.

More to the point, a team of more than 100 physicists at IBM had been contracted by the NSA to turn the Cryotron into the bedrock of American computing power. Project Lightning, as it was code-named, was the focal point of a new, top-secret meeting of advisory an committee on supercomputers that had been created by president Dwight D Eisenhower and of which Buck was a member.

When Buck learnt that the USSR’s top computer experts would get to breeze through his lab, he was left with a sick feeling. The trip had been arranged months in advance. He had noted the date, writing “RUSSIANS 2PM” in bold capitals in his diary.

He chewed over how to deal with the situation. There was little point in being too precious with information. 
A paper he had published four months earlier explained the experiments 
he was working on in considerable detail.

If the KGB – the Soviet intelligence service – was anywhere near as good as it was thought to be, then Lebedev would surely have been given a copy before his trip.

Soviet agents had been aware of Buck and his connections to the US secret services since at least 1952, based on declassified CIA files. It seemed reasonable to assume they knew who it was they were meeting.

The meeting

Lebedev and his team were more than a day late when they finally touched down in New York on Sunday April 19, 1959. Their chaperones – scientists from IBM and other military contractors – had been waiting in the terminal for hours. There had been no phone call or telegram to warn that they were no longer coming.

When their plane showed up without them, a check of the passenger records showed that they had missed a connecting KLM flight in Amsterdam. The following morning, however, word got through that their inbound plane from Moscow had been late the previous day, and they had missed their connection.

The lead host, IBM’s Mort Astrahan, and the assigned translator Eugene Zaitzeff, of Bendix Systems, would later coauthor an article for the Association for Computing Machinery’s Communications of the ACM that detailed the whole trip. A second translator had also been included in the group. Professor Lipman Bers, a Latvian-born mathematician from New York University, was also waiting in the terminal, along with his wife.

Bers had also been asked to act as translator for the corresponding American trip to the Soviet Union, the details of which had yet to be arranged. He and his wife had volunteered to act as hosts to the group of scientists while they stayed in New York, showing them a slice of what the Big Apple had to offer. Although the eight-strong delegation exchanged friendly handshakes with their hosts, the trip had taken its toll.

Their hosts bundled the group into cars and whisked them into town to their Manhattan hotel. Astrahan and Zaitzeff, both out-of-towners themselves, were staying in the same hotel as their Soviet guests. “Although the visitors must have been very tired, they then invited their hosts to a vodka and caviar party in one of their hotel rooms,” wrote Zaitzeff. “After friendly conversation a few arrangements were made for the last two days’ stay of the group in New York, which would include dinner at Professor Bers’ home, and also a Saturday trip through the stores on Fifth Avenue conducted by Mrs. Bers. Then the group broke up and each retired to his room.”

That Monday morning was wet and miserable. After breakfast, the first stop the group had to make was to the Park Avenue building where Soviet diplomats to the United Nations were based. They were given bankers’ drafts - their allowance for the two weeks they were in the United States - and told to cash them at the First National City Bank.

They went to the nearest branch, but were told that the drafts they had been given could only be cashed at the bank’s head office on Wall Street. Already the trip was taking on a farcical tone. As Zaitzeff wrote: “Since the group was expected in [IBM headquarters at] Poughkeepsie by 1:30 that day, time was running out. We decided to check out of the hotel and pick up the cars which were rented for the trip to Poughkeepsie.

"Somehow, in the confusion, the first cab took off without any English speaking individuals inside. Dr. Astrahan and myself found ourselves with the rest of the group in the second cab. Thus, upon arrival at the Hertz office and paying off the cabs we found that Mr. Glushkov forgot his briefcase inside the cab, not knowing that this was his final stop, and thinking that the cabs would be used to take us all the way to Poughkeepsie. Everyone became even more concerned when we were told that it contained three bottles of vodka and about six pounds of black caviar. It was never found.”

What the Soviets found

IBM had lined up some Russian speakers from among their ranks to join the welcome party. Some of them were wartime emigrants from Eastern Europe. Over two days they were shown the production lines for the IBM 704 and 705—the machines that were being installed in offices across America. Then they were taken to the third floor, where they saw the production of magnetic core memories—the RAM technology that Buck had helped invent.

They were even shown an early attempt at artificial intelligence: an IBM 704 machine was running an experimental program for playing checkers. Lebedev’s number two decided to take up the challenge: the machine conceded after only a handful of moves. On their car journeys between stops, Zaitzeff was bombarded with questions every step of the way. The group asked the population of each city they visited and were astonished when Zaitzeff told them he did not know, or gave only a rough approximation.

They were baffled to see the number of skyscrapers that were being built in New York—they had been told in the Soviet Union that America had concluded this was an inefficient way to build that always lost money. They grew a grudging respect for American cars, and asked a torrent of questions about the cost of each make and model they encountered on the road.

The answers “were not readily believed when the conversation involved used car lots,” wrote Zaitzeff. At 4:30 p.m on the second day the group was driven to LaGuardia Airport to get a flight to Boston. The next day they would meet Dudley Buck.

Upon arrival at MIT, Buck kept his guests waiting in the MIT staff canteen. After lunch the group was led up to the third floor of Building 10 and into Buck’s lab. The room was a hive of activity, as usual, but no one was making cryotrons or trying to build microchips.

Buck told them that the helium canisters needed to operate the cryotron were being refilled and thus a demonstration was impossible. Given that the visit had been in the diary for months, it seems unlikely that the Russians would have believed this excuse.

Ever patriotic, Buck clearly just didn’t want to show them his work. The Soviets made no complaint about the failure to see Buck’s work, even though he had been the headline speaker at a conference they had asked to attend – a diplomatic request that had precipitated the whole exercise that had brought them there.

They went on to Harvard, then to Philadelphia, Washington DC, and back to New York. Most nights ended after dinner with vodka and caviar provided by the Russians for their respective American hosts. They went to the cinema, where they complained about the “inferior direction”, and even learned to pay golf.

Just 29 days after the Soviets visited MIT, Dudley Buck was dead.

Douglas Buck
The Cryotron Files