From a small building in Pennsylvania to widespread usage across the world, we track the compelling story of one of the greatest technological innovations in history, setting the stage for the age of data science.
Ginette: “I’m Ginette.”
Curtis: “And I’m Curtis.”
Ginette: “And you are listening to Data Crunch.”
Curtis: “A podcast about how data and prediction shape our world.”
Ginette: “A Vault Analytics production.”
Ginette: “Today our story starts at a business building.”
Curtis: “The building is in Philadelphia, Pennsylvania, on Broad and Spring Garden Streets to be precise. Envision the late 1940s.”
Ginette: “You see a man absorbed in thought entering the building, and you decide to follow him in.”
Curtis: “When you walk through his office, you find some bright engineering minds working on a fairly new startup in town: the Eckert-Mauchly Computer Corporation, or EMCC. It turns out, this is the very first large-scale computer business in the United States.”
Ginette: “While this business environment on the surface is vibrant and innovative, behind the scenes, it’s a pressure cooker full of confusion.”
Curtis: “The owners, John Mauchly, who you followed into the office, and his business partner, J. Presper Eckert, are talking about something strange that’s been happening: most of their clients had been from the government, and now they’re quietly pulling away from doing business with EMCC without any explanation, which is both alarming and confusing to the business owners. It’d be one thing if the government gave a reason each time it pulled out of a contract, but without one, they have no idea what’s wrong or how to try and fix the situation. It’s like going through several breakups where the only explanation offered is, ‘it’s not you; it’s me.’
“So what’s actually going on here?”
Ginette: “The answer is woven into John’s backstory, a backstory that also includes the story of the ENIAC, the very first fully electric general purpose computer.
“In John’s earlier career, he was involved with scientific clubs and academia. He started as an engineer and eventually became a professor at the prestigious Moore School of Engineering at UPENN. At one point, he got lucky. He asked essentially this question to the right military person on campus: what if I could build a machine that would significantly reduce your trajectory calculation time for projectiles?”
Curtis: “So the military ends up formally accepting his proposal, and John and Presper team up for three years on this top-secret military project to build the ENIAC.
“At the time, the ENIAC is really impressive in both size and ability. It weighs about the same as nine adult elephants, which is 27 tons, and it has about 17,500 vacuum tubes, each about the size of your average household light bulb. It has 5,000,000 hand-melted joints. And it’s the size of a small house—about 1,800 square feet. And in today’s dollars, it costs about $7 million.
“It’s the very first of its kind. It’s both completely electric and a general purpose machine, meaning you can use it to calculate almost anything as long as you give it the right parameters. The bottom line is that it’s a lot faster than anything before it. It’s 2,400 times faster than human computers, and 1,000 times faster than any other type of machine computer at the time. For example, it took the calculation of a 60-second projectile down from 20 hours to just 30 seconds. To understand the magnitude of this, it’s like moving from an average snail’s pace to the average speed of a car on a highway.”
Ginette: “Here’s another way to look at this: if you drive your car (the ENIAC) across the country from L.A. to New York City at about 70 miles per hour without stopping, it would take you a little over a day and a half to drive there. In contrast, it’d take a snail (the human computer) without stopping about 11 years.”
Curtis: “So it turns out the ENIAC isn’t ready in time for aiding projectile trajectories in WWII—it’s not fully complete until February of 1946—but even before it’s completely ready, they could still use it to aid in calculations for the nuclear bomb. Once it is complete, it starts to create a buzz around the world because of its incredible speed, and it’s ability to maintain that speed because of its fully electric machinery.”
Ginette: “It’s about this time that UPENN plans to change its intellectual property policy—meaning John and Presper will not be able to keep their rights to their future invention, the EDVAC, which was an improvement on the ENIAC. So they leave UPENN, taking many people with them, and found this company, EMCC. John and Presper see that there’s a market in the business world for these computers. Many investors and businessmen don’t actually believe they’d be able to make a profitable business out of selling computers. Why? It’s mostly because they think there won’t be much demand for such a huge complicated machine. Businesses already have easy-to-use tabulating machines. Also, it took such an advanced understanding of math and logic to operate them that even if they mass produced them, there wouldn’t be enough people in the world who could operate them. In today’s context, it might be like trying to commercially sell space shuttles. Who’s going to buy them? Who knows how to actually use them? How reliable are they? What would businesses even use them for? In addition to those reasons, the machine is really hard to maintain. It has tons of vacuum tubes that would easily burn out and need to be replaced. We’re talking hourly maintenance.”
Curtis: “But contrary to this perspective, they actually do find a market for the computer. They target organizations that have to do extensive calculations, and they landed a deal with an organization swimming in unruly data: the US Census. The government employees in charge of conducting the census were under a ton of stress. The census was way behind schedule because of all the time-consuming calculations employees had to do, and they were basically drowning in literally a football field worth of population data on paperwork and punch cards. The stress of the situation goes beyond that, though. They absolutely had to get the work done by a drop-dead date because the elections are based on census data, and they had to be finished by then. And at the rate they’re going, there’s no way the census is going to meet that deadline.
“So EMCC wins a contract with the US Census to help them make their work more efficient, and if it weren’t for a little bit of misfortune, they probably would have succeeded. Unfortunately, this is where something that happened in Jon Mauchly’s past comes back to haunt him. Several years past, John had met with what he thought was a harmless scientific group. Some sources even say he only attended one of their meetings, but it turns out this scientific group had Soviet ties, and while the US government encouraged collaboration with the Soviets during WWII, the tides had definitely changed with the Cold War and past collaborations—which were once sanctioned—are now under intense suspicion. There’s a witch hunt for anyone who could have kind of communistic connections. Unfortunately, John Mauchly is one of these victims. Because of this one meeting he attended, he’s a suspected communist, and he loses his top-secret clearance with the government. Back then, the government didn’t have to give reasons why they’re pulling out of contracts.
“So after John Mauchly quietly loses his security clearance, the startup loses most of its contracts, and the company has to sell. They end up selling it to a big company, Remington Rand, and both John and Presper now have to go to work for this larger corporation.
“It eventually takes John ten years to clear his name of these false accusations, and by then of course it’s too late to save his company, and this left him bitter to the end.”
Ginette: “One thing this large company has that the startup didn’t have was killer marketing ideas—one scheme in particular permanently lodges the concept of the computer into the psyche of a skeptical culture. They strike a deal with CBS to showcase the computer’s predictive capabilities on television for the upcoming 1952 presidential election: Adlai Stevenson vs. Dwight D. Eisenhower. They think the machine can predict the president elect, and this is the first time a electronic computer has ever been asked to predict an election.
“So it’s election night, and the CBS news team gets ready to air the computer on TV, but before they broadcast the show, the computer, with only a sample size of 1 percent of the vote predicts Eisenhower will win by a huge landslide. The problem with this prediction is nobody on the CBS staff believes it. Everyone thinks the tally will be much more even. So chaos ensues on the set. What should they do? Should they really read such a blatantly wrong result on TV? If not, what do they say instead? They had to come up with a game plan fast, so they decided to tell the world, tongue-in-cheek, that the machine is just as baffled as the humans, and it’s really too soon to tell the results of the election. As the votes roll in over the night, it is clear that Eisenhower will take the White House by the largest landslide victory in the United State’s history. So, after midnight, CBS fesses up to the public and admits the truth. The machine had accurately predicted the election results. Everyone is stunned, and the marketing ploy works. After this stunt, computer starts popping up everywhere: the news, the movies, in cartoons.
“Here we start to see a shift of a greater number of people becoming aware of the computer, even though most people don’t have access to one. For that to happen, the computer had to shrink, a ton, and it did because of an incredibly important leap in technology. The invention of the transistor. The transistor replaces the vacuum tube, that piece that breaks down all the time in the computer and is the size of a lightbulb. Like the vacuum tube, the transistor controls the flow of electricity. This invention changes the entire field of electronics because everything becomes smaller and more reliable. The transistor is a 50th the size of vacuum tubes and weighs 100x less, and transistors barely give off any heat, which means smaller, faster, more complex, and more powerful computers.”
Curtis: “A transistor at the time is a very small structure with three-prongs that are coming out of it, and these three prongs have to be connected to other computer pieces by hand, which while reducing the size of the computer, it actually causes another really big issue issue—human error. Individuals have to hand-solder hundreds, thousands, and even tens of thousands of tiny metal pieces together, connecting everything through wiring. It gets more difficult as the designs become more complex and requires more electronic pieces and wiring to be connected. One slip up while melting these wires together can potentially make all the other hand-connected pieces completely useless.
“So, the amount of hand-wiring actually makes computers harder to make then they were before. And it’s such a big problem that they come up with a name for it, and they call it the tyranny of numbers. Until someone figures out how to overcome it, computers won’t get any smaller. Everyone recognizes this as the single most important problem in electronics, and everyone is trying to solve it. Think the late 1950s.”
Ginette: “With all this attention on this issue, there’s a major breakthrough: two men, who aren’t connected to each other in any way, one in Texas and one in California, create two very similar solutions to this massive problem. They both individually discover they can use one piece of material to create an entire circuit. This essentially combines a bunch of separate electronic pieces into one material, and eliminates wiring within it.”
Curtis: “They both come up with different designs, and the one that wins out is basically a piece of silicone chemically altered in different areas to create parts of the circuitry. Within the piece of silicone, metal is overlaid to connect these different pieces on the silicone. Because this circuitry on the silicon is connected together during the manufacturing process, it means a lot less hand wiring for computer builders. Now they don’t have to hand connect as many separate pieces because the wiring is a built-in part of the manufacturing process. Computer assemblers could basically just get the circuit delivered in the mail premade. But because it’s so expensive to produce, this invention doesn’t become huge until two years later in the 1960s when the space race pours immense amounts of money into it.”
Ginette: “Eventually, this solution, called an integrated circuit, or a ‘microchip,’ takes a whole board full of circuitry and shrinks it to a fingernail-sized chip. It’s orders of magnitude smaller, and it’s faster than anything the industry has ever used before.
“As microchips become mass manufactured and inventors find easier and easier ways to make them quickly, a $1,000 chip drops in cost to pennies and every one to two years, the microchip designs literally double in the amount of circuitry that they can hold.”
Curtis: “One incredibly significant spinoff from to this invention is really important to mention here for our story. In 1971, inventors shrink a whole computer microprocessor onto a single microchip. This means that something the size of your fingernail can now run all of the instructions for a computer program and perform all the calculations and logic for that program.”
Ginette: “Even with this invention that shrinks everything, people don’t have computers for their own personal use because basically there’s no reason for people to have a computer because there’s very little general expertise on how to even use one. Or so the computer industry thinks.”
Curtis: “In January 1975, a company advertises a kit you can buy to make your own personal computer in a tech hobby magazine. It’s called the Altair. It offers hobbyists a chance to build their very own personal computer. Nobody in the main computer industry, like IBM, thought that there was a strong market for selling computers to your average person. But by this time some people on the fringes of the computer industry had experienced the power of a computer at work or at school, and they wanted their own, if for no other reason than to have a modern marvel in their houses.”
Ginette: “So, people drive all night to buy this kit. This personal computer ad ignites a microcomputer revolution that had been quietly building, and tech enthusiast clubs start mushrooming up all over the country. These hobbyists get together in clubs to show off what they’ve done and how they’ve been adapting their personal computers. Small companies see a chance for profit here and start selling additional parts for the Altair so people can improve their microcomputers. And soon small companies are building entire personal computers.
“A little over a year after the Altair goes up for sale, these tech enthusiasts host a microcomputer convention in Atlantic City, New Jersey. Here they show off their souped up computers. It turns out, three young men are there, one of whom is described as a quote ‘renegade from the human race.’ This person turns out to be Steve Jobs, the man of legend. These young men eventually find themselves at the forefront of a microcomputer revolution, and with the help of an experienced mentor who used to work for Intel, Apple computers is born.
“Apple initially dominates. Everyone wants a bite of this fruit, and the company makes its initial employees and investors incredibly rich. In fact no previous company in history had made so many so rich so quickly. And the rest is a familiar history, so we’ll end this episode by speeding forward to the 2000s and making a few interesting comparisons for you.”
Curtis: “Let’s compare the ENIAC to your average 2007 cell phone. Going back to the original speed comparison that we made earlier, we’re taking a trip from L.A. to NYC. We already know a human computer is essentially a snail that takes about 11 years to get there, and the ENIAC is the car going 70 MPH and takes about a day and a half. Let’s look at how fast a speeding bullet would make the journey, assuming a constant speed and no interference—it would take about an hour. With that as a marker, to compare the ENIAC to the cell phone, we actually need to go about 35 times faster than the speeding bullet—making the trip in about two minutes. That’s faster than the space shuttle and just under the top speed of the fastest man-made objects ever, which are the Helios probes that NASA sent to the sun.
“Just for fun here are a few comparisons that will illustrate the relative differences between the ENIAC and a cell phone using everyday objects. So comparing the power consumption of the ENIAC to the cell phone is like comparing the City of Chicago’s power consumption to turning on two toaster’s in your kitchen.
“Comparing the size of the ENIAC to the size of the cell phone is like comparing the Burj Khalifa and a gallon of milk.
“For the difference in price, consider the difference between your average new four door sedan and the price of a Snickers bar.
“And to consider the weight difference, consider the difference of a blue whale and a loaf of bread.”
Ginette: “And what about the transistor? It’s mind boggling to think that something that replaced a computer part the size of a lightbulb, we now measure its size in terms of atoms.”
Curtis: “So now we’ve now arrived in modern times. In this three-part series, we’ve gone from hand-created data graphics to the development of the modern-day computer. We’ve set the stage to understand how our day is completely unique in the history of humanity. Coming up, we’ll dive into a job occupation that has emerged as a result of all of these technologies and ideas coming together—known today as data science. We’ll explore how this new discipline is and will change our world in ways that we can hardly imagine.”
Curtis: “Before we close out the show today, we want your feedback. We’re looking for the most most interesting and unexpected ways that data has affected people’s lives. So if you have a story, then we’d love to hear about it. Send us a message on Twitter at @vaultanalytics or on our website contact-us page, at vaultanalytics.com. And we’ll be reading all of the feedback and mentioning some of the most interesting stories on a future podcast episodes.”
Some Sources:
Media,Technology and Society: A History: From the Telegraph to the Internet
Music Sources
Public Domain music from FreePD.com.
“Cold Funk” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 3.0 License
http://creativecommons.org/licenses/by/3.0/