More on this book
Kindle Notes & Highlights
Read between
October 24 - November 2, 2025
Picture the communications network as a series of nodes: If there is just one connection between two nodes and it is destroyed in a nuclear attack, it is no longer possible to communicate. Now imagine nodes with multiple connections to other nodes, providing an alternate path of communication if some nodes are taken out. The question for Baran was, how much redundancy is enough? Through simulations of an attack, Baran and his colleagues found that if you have three levels of redundancy, the probability that two nodes in the network could survive a nuclear attack was extremely high.
Pentagon officials did not quite understand what Licklider was talking about, but it sounded interesting, and Ruina agreed, or at least he agreed that Licklider was smart, and the specifics were not important.
Ruina, an engineer, was even less interested in the behavioral sciences, Licklider’s other assignment, which was allocated just $2 million a year. Ruina dismissed the entire field as Freudian ruminations. “Tell me what has happened in the last twenty years in behavioral sciences that you would think of as a breakthrough in the sense of giving us new concepts, and thinking, and important contributions, and…did it come from any government contract—cut-pipe work, or was it a guy who is more of a novelist like Tolstoy who was able to do great human insights without having to get a government
...more
Licklider ended up spending most of the behavioral science money on human-computer interaction, rather than anything related to social science, which suited Ruina just fine.
Newcomers like Licklider were essentially making up the rules as they went along, creating what would later be regarded as the hallmark of ARPA: freewheeling program managers given broad berth to establish research programs that might be tied only tangentially to a larger Pentagon goal.
Licklider, who focused on vision over reputation, also took a risk on more unknown scientists, like Doug Engelbart, at the Stanford Research Institute. By the time Licklider was done handing out contracts, his centers of excellence stretched from the East Coast to the West Coast and included MIT, Berkeley, Stanford, the Stanford Research Institute, Carnegie Tech, Rand, and System Development Corporation.
“At this extreme, the problem is essentially the one discussed by science fiction writers: ‘How do you get communications started among totally uncorrelated ‘sapient’ beings?” he wrote. The six-page memo went on to state explicitly what he had in mind. “It seems to me interesting and important nevertheless to develop a capability for integrated network operation,”
When Ruina left in 1963, his replacement, Robert Sproull, a scientist from Cornell University in Ithaca, New York, almost canceled Licklider’s entire program.
Asked decades later about whether he was the man who “almost killed the Internet,” Sproull laughed and said, “Yes.”
At MIT, the ARPA-sponsored time-sharing system spawned the first e-mail program, called MAIL, written by a student named Tom Van Vleck.
Ivan Sutherland, a brilliant young computer scientist who had already forged an impressive reputation for his work in computer graphics, replaced Licklider. But at the time he was brought into ARPA, Sutherland was just a twenty-six-year-old army lieutenant, recruited because no one else qualified wanted the job. ARPA’s unusual system of having only temporary employees made the position difficult to fill. Government salaries were low, and there were no provisions at that point for temporary academic appointments.
Sutherland, however, had no choice. “I was in the army, and I got some orders which said, ‘You are hereby ordered to proceed to the Pentagon and take this job,’ ” he recalled.
University professors were scared that networking computers would allow others to tap into their coveted computer resources.
Taylor was not proposing Licklider’s small-scale laboratory experiments of a couple years earlier. Taylor wanted to create an actual cross-country computer network—something that had never been done before and would require significant new technology, investment, and arm-twisting of researchers.
The conversation to approve the money for the ARPANET, the computer network that would eventually become the Internet, took just fifteen minutes.
The Internet might have no single progenitor, but Licklider’s ability to carry out his vision from Washington, while Baran’s perfectly good idea died on the vine at Rand, underscored ARPA’s unique position in the 1960s. The ARPANET was a product of that extraordinary confluence of factors at the agency in the early 1960s: the focus on important but loosely defined military problems, freedom to address those problems from the broadest possible perspective, and, crucially, an extraordinary research manager whose solution, while relevant to the military problem, extended beyond the narrow
...more
If you are not careful you can con yourself into believing that you did the most important part.
In later years, Ruina was interviewed frequently about his role in the development of the Internet. He often repeated some variation of the same theme: he really had little idea what Licklider was doing, other than he thought it was something good.
The extraordinary freedom given to ARPA in the early 1960s to tackle military problems was supporting fruitful research into computer networking and nuclear test detection, but in Southeast Asia the agency was pursuing something equally ambitious, but with far darker consequences.
One of the first and most significant of those experiments was chemical defoliation, which was under the personal control of William Godel.
The bands on the barrels of those first aircraft were purple, denoting “Agent Purple,” a fifty-fifty mixture of two herbicides: dichlorophenoxyacetic acid and trichlorophenoxyacetic acid. Other agents included Pink, Green, Blue, White, and what would eventually be the most widely used defoliant in Vietnam, Agent Orange.
All of Godel’s AGILE projects were tied together. Chemical defoliation would cut off food sources and eliminate ground cover for the Vietcong in the jungle. The South Vietnamese military would be provided, courtesy of ARPA, a host of new technologies—such as new lightweight guns—to fight the Vietcong in the jungles. Finally, the linchpin of the strategy was the strategic hamlet program, which would resettle farmers and their families to areas that could be secured, cutting off opportunities for the Vietcong to terrorize, recruit, and resupply. Or so went the plan.
In Malaya, the communists were ethnic Chinese, a distinct group from the Malay, so when the population was resettled, it was relatively easy to keep the Chinese out of the new villages. In Vietnam, however, there was no easy way of distinguishing the Vietcong from peasants
Hickey’s report to ARPA on the strategic hamlet program, which was resoundingly negative, was not greeted with enthusiasm in Washington, either from civilian or from military leaders. Harold Brown, one of many Pentagon officials Hickey briefed, literally turned his back on Hickey and a colleague during their briefing, swiveling his chair away from them.
Unable to adapt to the jungle, Godel’s dogs had largely grown sick and died or were eaten.
At one point, Colonel Trach, the Vietnamese head of the Combat Development and Test Center, started selling the barrels, only to get in trouble, not because the barrels were dangerous, but because he failed to give a cut of the profits to his higher-ups.
Yet when advocates of the AR-15 back in Washington used the positive reports to try to persuade the Pentagon to buy the new rifle for American soldiers, it ignited a political firestorm among the White House, Congress, and the Pentagon. The army did not want to be told what weapon to buy, and its leadership believed that the AR-15 was not as lethal as the M14.
Godel had wanted the AR-15 to help the Vietnamese soldiers battle an insurgency and instead saw his idea get bogged down over an entirely different question: the fielding of a new weapon for American troops.
ARPA reported not to the commanders in charge of operations in Vietnam but instead to Harold Brown back in the Pentagon. According to an official army history, the military services looked at ARPA as an unwanted competitor, and they “regarded the ARPA field unit with distrust.”
Agents turned their attention to Godel, who was used to spending cash on covert operations with few or no questions asked. In the Pentagon’s Office of Special Operations, Godel’s penchant for spending money on covert operations without proper approvals had gotten him reprimanded, but nothing serious. Now he was at the center of an FBI investigation.
Not present at that dinner was William Godel, who within months would be fired from his position at ARPA and indicted on criminal charges of fraud.
Wylie was found guilty of embezzlement and fraud. Godel was acquitted of embezzlement but found guilty of making false statements and conspiring with Wylie. For Godel’s family and friends, it made no sense that the jury would clear him of stealing money and yet find him guilty of working with Wylie, a man Godel could barely tolerate. Both men were sentenced to five years in prison.
Asked ten years later in an interview commissioned by ARPA to list any true successes from AGILE, Godel gave a one-word answer: “None.”
As for the conviction of his onetime mentor, the only thing Herzfeld would say on the subject was “stuff happens.” Herzfeld’s vision for ARPA, and AGILE, was more expansive than any previous director’s. For Herzfeld, the world was a giant laboratory.
“Is there any practical way to simulate the behavior of a nation in a laboratory?” Mahon, in apparent disbelief, asked ARPA’s director, Robert Sproull. “I believe so,” Sproull answered confidently.
Herzfeld carried himself with the confidence of a senior statesman. He embodied ARPA at its height: when it succeeded, as it did with the ARPANET, it succeeded big and helped change the world. But when it failed, it failed big, and that changed the world, too, and not always for the better.
His first assignment was to write a political report on Latin America. He rushed out and bought a book on Latin America by Lincoln Gordon, the American ambassador to Brazil. “I plagiarized a lot of stuff,” he admitted later.
Stark went with another ARPA official to a local hostess bar, where a Thai woman’s company could be bought by the hour. He paid for an hour, not out of romantic interest, but so that he could have someone who would tell him something about Thai culture, about which he knew nothing.
the agency ultimately envisioned creating a global database of people, politics, and places that could then be applied as a model anywhere in the world where the United States needed to battle insurgency.
Fighting guerrillas in the mountains? ARPA would know what equipment would be the most effective at high altitudes. Need to create a “hearts and minds” campaign for a region infiltrated by pro-Castro communists? There again, ARPA could draw on its database of social movements.
One example was a “sunlight projector” developed by ARPA for psychological operations teams. As the name implied, it was a movie projector that used a mirror—a shaving mirror in this case—to focus sunlight on movie film to project it on a screen. The idea was that this sort of ruggedized equipment could be easily carried into small villages to play government propaganda films. An American psychological operations team actually tested the device in 1965 in a small village. The results were not good: the projector ended up focusing too much sunlight and burned a hole through the film.
As Stark walked down the street in Thailand, Thai children would often call him farang kee nok, or “bird shit foreigner.” Stark would answer with farang kee Tai, or “bird shit Thai.”
Murray Gell-Mann, a physicist, whimsically suggested ARPA look at the effects of various security tactics, like cutting off the ears of insurgents.
Just a couple years later, Gell-Mann won the Nobel Prize for his theory of quarks, confirming that brilliance in physics does not necessarily translate into other fields.
“You helpless creature,” Khomeini said, addressing the shah directly, “you don’t realize that on the day when a true outburst occurs, not one of these so-called friends of yours will want to know you.”
In 1942, H. Norman Schwarzkopf Sr., chief of the New Jersey State Police, was sent to Iran to help to remake the gendarmerie “in the mold of the New Jersey State Police.” Schwarzkopf, whose son would lead American forces in 1991’s Operation Desert Storm, was credited with remaking the gendarmerie into an effective counterinsurgency force that put down tribal rebellions in Kurdistan and Iranian Azerbaijan.
Eventually, however, ARPA was told by the Iranian government to stop its work on detecting heroin smuggling because it had “gotten too close to the top level traffickers,” Herzfeld recounted in his memoir. “That ended that.” The problem with the ARPA counter-smuggling program, like almost all of the agency’s work in Iran, was that it presumed the Iranian government wanted to solve the same problems the United States wanted to solve.
“I guess one way you look at it: What might have been the sources of future insurgencies. Which families?”
It was not that the two projects were inherently contradictory but simply that ARPA did not seem to have any larger goal other than planting its flag in the region,
One proposal, for example, suggested using chemical weapons to protect the president. “There also exists a need for a system which would make an unfriendly crowd become friendly almost instantaneously,” one note in the files stated. “This goes beyond the desire to divert a crowd, as could be done by the prompt and generous use of cash money. The possible use of gasses, sound, lights and other chemical biological or psychological agents to achieve such a change as well as other attributes they might possess for crowd control will require further study.”
