Averting the Apocalypse with Hot Potatoes: How Paul Baran Invented the Internet

univac_i_interior.jpg, default, Internal view of UNIVAC I, photo: Alejandro Quintanar / Wikimedia Commons / CC BY-SA 4.0

#technology & innovation

#heritage

#digital culture

Ever wondered how it is that whenever you click on one of Culture.pl’s links, you’re immediately transported to a different place, full of wonderful pictures and well-written sentences telling a fascinating story? Well, it was all made possible thanks to the work of Paul Baran, a Polish Jewish American engineer who laid down the Internet’s technical foundation.

First, a thought experiment

Thumbnail — Paul Baran, photo: Wikimedia Commons

Imagine you had to deliver a package with some bits and pieces in it. It has an addressee, but nobody really knows how to reach them. The only available information is that they’re part of the delivery network and, surely, somebody must be familiar with them. How would you go about delivering the package? The simplest solution, really, is to just ask people to pass it along, hoping that it’ll eventually reach its destination. If everybody is in close contact with a few people and the package is returned to the previous holder only if there are no other options, there are genuinely high chances that the box of goodies will finally get delivered.

Now let’s take it further and imagine that those bits and pieces I mentioned are computer bits encoding information that can be replicated countless times and travel extremely fast. Following the same method, all you’d need to do is make sure that every computer in the network was connected to at least one other device. You encode the address on the packet of bits you want to send (or even its multiple copies) and get it on its way, with each computer passing along anything that is not addressed directly to them. Sooner or later, this digital game of ‘hot potato’ results in the package ending up in the right hands, especially since it has a counter listing the number of steps taken and is prompting the network to look for the shortest possible route.

However weird this method might sound, it actually works and fuels the Internet as we know it – the moment you clicked on the link to this story, a server somewhere in the world constructed a package containing all the zeroes and ones needed to tell your computer to display this webpage to you. After some exchanges, the hot potato eventually reached its target, except in this twist on the playground game, you actually get rewarded by holding onto it.

Picture display

standardowy (864px desktop)

Exterior of Benjamin Franklin Institute, Philadelphia, PA, photo: Joe Sohm / Getty Images

Picture image

gettyimages-73071934.jpg

The ‘hot potato heuristic’ is a really helpful way to illustrate what we today call ‘packet switching’, because it was created for this very reason by the technology’s inventor – Paul Baran, a Polish Jewish American engineer, who can quite rightly be called ‘the father of the Internet’. Although the credit for creating the World Wide Web usually goes to Tim Berners-Lee, who allowed us to get online by inventing the rules upon which the Internet operates, none of it would have happened without the introduction of packet switching.

To use simpler terms, Berners-Lee showed us how to address letters and where to glue down stamps. He made it so easy that today everybody can use the postal service. But he could never have done this without Baran, the person who first found a way how to actually get those letters to their destination (interestingly, he also liked to use the postal metaphor). The thing is, the hero of our story did not exactly plan any of this.

The source

Picture display

standardowy (864px desktop)

Grodno, a general view of the city and the Niemen river with a wooden bridge in 1920 from the Łosośna neighbourhood, photo: NAC

Picture image

grodno_1920_-_widok_miasta_i_rzeki_niemen_z_mostem_drewnianym_nac_1-u-1395-2.jpg

Let’s trace everything back to its source. Paul Baran was born on 29th April 1926 in Grodno as the second child of Morris and Anna Baran. However, the young Psachija, as the boy was first called, didn’t get to spend much time in the city of his birth. Between 1914 and 1920 – during the turmoil of First World War, the subsequent eventful years and the Polish-Soviet war of 1920 – Grodno moved across different countries numerous times. Unsurprisingly, Morris and Anna decided that it wasn’t exactly the best place to raise children. They applied for a visa and in 1928 arrived at Ellis Island, New York (Paul Baran later wrote that he moved to the US at the age of two and that it was one of his best decisions). They were greeted by their relatives who helped them first get to Boston and later settle down in Philadelphia, where Morris opened a grocery store.

The first memories of young Paul (his and his siblings’ names were anglicised upon arrival to the US) must have been of West Philadelphia, where he grew up and went to school. His old neighbourhood does not have a good reputation today, but at the time it was a bustling suburb and gave Paul good prospects for the future – he attended a decent high school and could easily visit the Franklin Institute, where he first became fascinated with natural science and technology.

Picture display

standardowy (864px desktop)

Drexel University, main building, photo: drexel.edu

Picture image

drexel-main-building.jpg

Even though he loved learning, he was not your typical bookworm. His interests were pragmatic, as demonstrated in his early teens when he started his first business-like venture – he turned his starting capital of pennies into ‘dimes’ by coating them with mercury from a chemistry set. Around the same time, he also developed an interest in HAM radio (his call sign was W3KAS), which later influenced at least some parts of his career.

Baran later enrolled at a co-operative course at Drexler University and paid his tuition by working for several local radio stations by servicing their equipment. He graduated as a Bachelor of Science in Electrical Engineering in 1949 and went on to start his first full-time job, which was already an interesting fit. He became an engineer working for Eckert Mauchly Computer Corporation (EMCC), most famous for creating the Univac: the world’s first commercially available computer.

This thing has no future

Picture display

standardowy (864px desktop)

U.S. Air Force technicians evaluate the UNIVAC computer system in 1951 which took up 352 square feet of floor space and ran at a then-astronomical rate of 2.25 megahertz, photo by U.S. Air Force / Getty Images

Picture image

gettyimages-696806.jpg

While this position might have looked quite impressive on young Baran’s CV, he never actually got to see a working Univac. Shortly after joining the company, the Polish American engineer decided to find out what this new exciting field of informatics was all about and attended a course at the University of Pennsylvania. Unfortunately, the rows of zeroes and ones used to represent information were so incomprehensible to him that he found the whole thing absurd. He decided instead to focus on the technical side and went on to test the highly-unreliable vacuum tubes that the Univac used to store inputs.

Having engineers as skilled as Baran working on various aspects of the new revolutionary digital device should have been sufficient for the company to succeed in creating the first programmable computer. However, the ground-breaking technology created by EMCC looked quite different to what you’re probably using to read this article. Univac measured 7.5 by 15 metres and cost around a million dollars (10 million today, adjusted for inflation). This is not a price point at which customers jump at to try new unproven devices, so unsurprisingly, EMCC ran into deep financial troubles and until their first investor came along in 1951, the chances that they would be able to successfully build the machine they had on paper were pretty slim.

This is why Baran quit EMCC in 1950, having decided that even if the company managed to sell a dozen Univacs, their business model was not exactly sustainable. As he saw it, there simply wasn’t enough demand for these complex machines (keep in mind that in order to make full use of a computer, companies needed to change their whole way of operating, which incurred additional cost). Even though he later admitted that his decision was premature (Univacs became quite popular: one famously predicted Eisenhower’s win in the 1952 presidential election and another operated in Tennessee until 1970), Baran decided to find a new employer and focus on a technology that in his eyes actually had a future: radio.

Radio days

Today’s action movies have made us take ballistic missiles for granted. The bad guys only need to press a single button to destroy half the planet, as the hero bravely steers their plane to avoid incoming self-guided rockets. But this was far from the case in 1950. Back then, the only way to hit an enemy with a missile was to shoot a bunch of them in the target’s general direction and hope at least one of them made it. And this is precisely what Baran’s new employer, Raymond Rosen Engineering Products set out to change.

In October 1950, they created a revolutionary technology for the military, which allowed operators to transmit and precisely record the in-flight measurements of a missile’s trajectory. This in turn, made it possible to guide the missile directly, offering the military a degree of accuracy unheard of during the recently finished Second World War.

Except, of course, it wasn’t that simple. Baran’s employer did create the system, but not everything worked perfectly. The signal was recorded on magnetic tape and any attempts to read it resulted in huge amounts of static, which undermined the recording’s very purpose. Baran’s job was to find a reason for the interferences and fix it as soon as possible – the military was certainly interested in the company’s invention, but competitors were working on their own missile-guidance systems.

Picture display

standardowy (864px desktop)

Martin Matador cruise missile being launched from a rail with rocket assist, photo: public domain

Picture image

martin_mgm-1_matador.jpg

Baran realised that the magnetic tape itself was an issue. It was a relatively new technology at the time and the surface of commercially available tapes was slightly uneven, which resulted in static – nothing horrible when you’re only trying to listen to Nat King Cole’s hot new single, but you probably wouldn’t’t want to guide missiles this way. The young engineer quickly ordered specially-designed perfectly-even tape, and together with his colleagues developed a new electronic static-reduction device. With this impressive contribution, he was invited as an expert to help out with the MGM-1 Matador missile’s test launch in the Bahamas. As Baran himself later recounted, the missile flew up into the sky and started descending, ready to hit a small island and display American military prowess to all the spectating officers. Too bad it missed its target.

Baran quickly started analysing his tapes and found out what the problem was. But he didn’t give his superiors the solution outright, instead opting to revel in his moment in the spotlight because the military personnel had largely ignored him before as a mere technician. In the end, he and his colleagues demonstrated to the brass that the system did not fail. It’s just that the island was not where it was supposed to be: the military was depending on imprecise maps. The second launch was a success (although it has to be noted that this anecdote is nowhere to be found in the official report) and the military was then very eager to work with Raymond Rosen Engineering Products. The company only needed to sign the papers, but the owner decided that his traditional business of fixing the radios and TVs of local customers was much more reliable than some finicky government contract.

Probably still feeling excitement from the Matador launch, Baran quit his job. The next few years were quite eventful. Working for Ampex, he developed a new method of radio transmission which was resistant to interference. He met and married Evelyn Murphy, the love of his life (apparently, he won her over with his sense of humour and honesty – when Evelyn ruined the dinner she prepared for their first date, Paul asked: ‘Are we eating this or is it already eaten?’). He moved to Los Angeles to start a new job at Hughes Aircraft where he worked on a transistor-based digital radio signal processing system that was a part of the ground-breaking SAGE computer command system. And since his newest job meant he was involved with computers again, he decided to get a Masters degree in informatics from UCLA. Not a bad way to spend your late twenties and early thirties.

But his biggest challenge was waiting for him in the RAND corporation.

How to build a strong network

Picture display

standardowy (864px desktop)

UCLA campus in Los Angeles, California, photo: Getty Images

Picture image

ucla-campus-gettyimages-166041568.jpg

The famed RAND was created to solve the most vital problems faced by the US government. And when Baran started working there in 1959, the US government wanted to know only one thing: how do we win the Cold War?

As you may recall, Baran had already contributed to the Cold War effort when he helped developed the missile-guidance system powering the Matador nuclear warheads. But, as experts in national security will tell you, nobody really wants to use their nuclear arms – it would mean game over not just for the enemy, but for life on Earth as we know it. The US had the means to obliterate at least the major Soviet cities, but the technology was not one of a kind – the Soviets also had their own intercontinental ballistic missiles and were just as able to lay waste to the Americans and their allies. It was only a matter of who was going to be the first to push the button.

RAND was tasked with ensuring a second strike capability, but even if the military wanted a way to hit the enemy back, the institute’s employees saw it as a potential deterrent. If the enemy knows that the US is able to launch their missiles even if their central bases are destroyed, they might be less inclined to start a nuclear apocalypse, even if they were really eager to do so. All that was needed was a communications system that worked after nuclear blasts had obliterated a significant percentage of its nodes.

Picture display

standardowy (864px desktop)

Centralized vs distributed networks, photo: historyofcomputercommunications.info

Picture image

ex2.0_centralized_vs_distributed_networks_photo_historyofcomputercommunications-info.jpg

Baran quickly got involved in the project since he had seen how dangerous guided nuclear missiles could be and he wanted to make sure that superpowers had as few reasons to use them as possible. He started experimenting and came out with the idea of creating a system of cheap unreliable transmitters with extremely high redundancy. This would mean that whenever one element was destroyed or did not work for some other reason, another would be able to pick up its task – in this case, transmit a message. It turned out that in order to guarantee the US a chance of responding to a nuclear strike, every node in the network had to be connected to just five others. Even if the majority were destroyed, messages could freely move between the remaining elements.

The engineer proposed using digital means (other than computers!) to send messages of 1024 bits, which equals 128 signs. Coupled with the address of the recipient and the number designating the packet’s order within a bigger message that was being relayed, the system made it possible, with enough patience, to relay any information imaginable across the entire United States. Baran wittily noted in his publication on this subject that the creation of the system would not even require raising taxes and compared the cost to the first mobile telegraph created during the American Civil War (discussing the telegraph, the engineer factored in even the costs of the oats for the horses involved in the construction, because he believed ‘in rewarding good work’). Now somebody only needed to build the whole thing.

Playing telephone

At the beginning of this article I wrote that Baran can quite rightly be called ‘the father of the Internet’. This is not an uncontroversial statement as some Internet historians credit Donald Davies, a British computer scientist, with inventing the ‘packet switching’ for which Paul Baran is the most famous. In fact, Davies definitely invented the term, which Baran considered very apt, and he also independently conceived of the entire process a few years after the RAND engineer had already outlined its foundation. Even though there is some debate whether he had heard about Baran’s ideas or was somehow inspired by them (for example, Baran remembers chairing a conference in which Davies participated), the military network ARPANET, which was later split into a public Internet, was created on the basis of Davies’ work and Baran only served as consultant.

However, as Baran’s biographer Wojciech Orliński aptly notes, ‘if – following patent law – we take an inventor to be the person who first described an unobvious invention in a way that would allow a person versatile in the craft to make it, then in the case of the Internet, Paul Baran would unquestionably be its inventor.’ Baran not only described the Internet as we know it, but he also almost built it – the entire project never got off the ground though because the Polish American engineer sabotaged the whole thing.

Following his ground-breaking work on packet switching, Baran quickly convinced his bosses at RAND to start development. This was a feat in itself, because the institute’s own employees, frustrated when their brilliant ideas were not brought to life, often famously deciphered the acronym as “Research And No Development”. The military jumped onboard very quickly – they were excited about an indestructible system for sending commands in the middle of a nuclear war and the fact that Baran’s main motivations arose from anti-war sentiments did not bother them as long as everything looked good on paper. The only involved party that was not enthusiastic about the idea was the telecommunications giant AT&T.

AT&T commercial from the 1960s

It’s not difficult to see why the company was sceptical about the project they were supposed to help develop (a mountain of paperwork and bureaucratic protocols made it necessary for RAND to work together with AT&T at the time). A cheap, reliable, long-distance, low-latency communications system was clearly direct competition to the telephones that were at the time still the foundation of the company’s business, even if, as Baran argued, telephones probably wouldn’t work after a nuclear strike. Moreover, Baran designed digital devices that were supposed to do all the communicating and ‘digital’ was not a word that excited the decision-makers at AT&T, which still used mainly analogue means of communication and constructed their network in a way opposite to what Baran proposed. It’s the equivalent of gifting somebody a Spotify subscription whose first reaction is to try make it work on a gramophone. This is how the engineer must have felt while explaining his idea.

The problem was that despite AT&T’s reluctance, the project ended up with them anyway. The military got so excited about Baran’s network that they ultimately tasked the Defense Communications Agency (DCA) with bringing it to life. Baran believed that the people in DCA would be just as unreceptive to his novel ideas and just as dependent on the old analogue ways of thinking as AT&T. Moreover, the leaders of the DCA had strong connections with AT&T and it was very likely that the company would be in some way involved in the planning and development of the network. Baran was afraid that not only would his plan fail under the supervision of incompetent engineers, but also that they would want it to fail. He contacted his friend in the Department of Defense who was responsible for allocating resources to similar projects and told him that creating the network in this way would be a waste of money: something that would never work. It seems possible that the Internet exists in its current form only because its inventor worked hard to slow down its arrival.

Unwittingly, Baran prepared the ground for people like Davies who used the Polish American engineer’s technical foundation and found a way to circumvent the problems which Baran had encountered – they used computers, taking AT&T out of the picture. In 1969, ARPANET, the Internet’s granddaddy, was created. On 29th October 1969, the first packet almost relayed the message ‘login’ from Stanford to UCLA. Almost, because one of the computers crashed in the process and only managed to send the letters ‘lo’. But the test was a resounding success – one hour later, the whole packet was delivered. Over half a century later, Paul Baran’s story has been delivered to you using exactly the same method.

Going offline

Picture display

standardowy (864px desktop)

Paul Baran with his wife Evelyn, photo: private archive

Picture image

paul-baran-z-zona-evelyn-archiwum-prywatne.jpg

Although Baran is most famous for packet switching, his life as an inventor would’ve been quite significant even if he had not contributed to the creation of the Internet. The list of his inventions includes a doorway gun detector (which later gave Baran headaches every time he boarded a plane), the first satellite and wireless Internet connection systems, an electronic remote electricity meter, the fastest (at the time) Internet modem, a portable printer, and many other pioneering things. In his later life, Baran created one technology company after another (today, we would call them ‘start-ups’) only to later sell them for millions of dollars and focus on his newest brilliant idea. He briefly retired, but ultimately decided that an idle life in sunny California was not for him and he came back to work on developing new revolutionary devices.

As impressive as this list may be, Baran was more than just an inventor – he was a visionary. Mindful of his youthful underestimation of the potential of computers, he later looked into the future with remarkable accuracy (after he left RAND, he was even one of the founders of the Institute for the Future, a think-tank specialising in technology foresight). Even though his networking project did not involve computers and ultimately did not come to life, he knew that digital communication would be at the centre of things to come. People wanted to be more connected and he had the skills and opportunities to make that happen (although, as Orliński notes, he depended a bit too much on equipping his devices with dedicated terminals instead of creating things that can be connected to a universal personal computer).

Baran’s predictions are even more striking than the things he built. Despite his battles and disenchantment with AT&T, he was aware that the future of communications lay with big companies like them, since they would be able to supply their consumers with TV, telephone and Internet using a single cable. In 1967, he gave a presentation (later published by RAND) in which he outlined how the world of marketing would change in the year 2000. No longer would people just walk to a department store and buy things on impulse, but they would use a terminal connected to a TV to browse a virtually infinite online catalogue, looking only for the exact thing they needed (in Baran’s eyes, an example consumer in the year 2000 would probably be buying a sabre saw), similarly to academic researchers browsing papers to find one dealing with a particular problem. At the time, his idea was met with opposition from professional marketers, but history ultimately proved him right – online giants like Amazon popped up in recent decades as if they had been following a business plan outlined by Baran.

Commercial for one of Baran's inventions, a wireless Ricochet modem

Perhaps most importantly, Baran saw the work of an engineer as something that was about more than just the creation of new devices and saving or making money for their employer. He argued that engineers have a social responsibility, because their creations are able to change people’s way of life in the future, especially as more and more technologies were becoming connected to entire systems with vast capabilities – he believed that these changes should only be brought about in a responsible way and not just because we can.

Shortly after laying down the foundations of the Internet, Baran turned to investigating a problem he saw as most pressing in light of the upcoming digital revolution: privacy. Back in 1968, he published an article explaining how the collection of personal data threatens individual autonomy and can bring societal problems that might be impossible to overcome (displaying his trademark wit, Baran also joked, for example, that the availability of dental records would betray those of our friends whose teeth were not their own). His answer to the problem involved more than just concentrating on practical solutions (although he was active on that front as well, and in 1965 he invented an end-to-end message encryption method which led to the creation of today’s encrypted communicators like Signal and WhatsApp). He believed that engineers should take responsibility for their inventions and address privacy and other ethical and social concerns at the level of design, ensuring that new technologies did not bring unplanned problems, but made them preventable.

In light of countless privacy breaches, online dangers and a general lack of responsibility amongst Internet giants, it would be best if we recognised Paul Baran not only as ‘the father of the Internet’ but also as a pioneer of digital ethics, advocating for the social accountability of technology-makers and the responsible design and development of technology. These are ideas which remain especially relevant today. Only then would the packet sent from Grodno in Poland in 1926 deliver its complete message.

Written by Michał Wieczorek, August 2020

_{Sources: ‘Człowiek, Który Wynalazł Internet: Biografia Paula Barana’ (The Man Who Invented the Internet: A Biography of Paul Baran) by Wojciech Orliński, 2019; ‘Some Changes in Information Technology Affecting Marketing in the Year 2000’ by Paul Baran, 1969; ‘On the Engineer’s Responsibility in Protecting Privacy’ by Paul Baran, 1968; ‘Conversations with a Pioneer: Paul Baran in His Own Words’ by Morten Bay, 2017; ‘Hot Potatoes and Postmen: How Packet Switching Became ARPANET’s Greatest Legacy’ by Morten Bay, 2019; ‘An Interview with Paul Baran’ by Judy O’Neill, 1990; ‘An Evening with Paul Baran, in Conversation with Henry Lowood’, recorded 2005, published 2011.}

Language English