by Brian Dear
Each time a student entered an answer to a question, the system immediately responded with OK, NO, or a customized message, depending on the correctness of the answer.
Another aspect of PLATO visible right from the beginning is what today’s Internet industry calls “analytics.” PLATO collected student data while students worked on lessons, and that data was then available for analysis by the instructor. A late 1960 progress report on PLATO had this to say:
The machine keeps an accurate record of each “move” the student makes. Thus at the end of an instruction period, the experimenter has at his disposal a print-out of how long the student spent on each page, what sequence of right and wrong answers were given, how long a problem required solution, as well as at what point help was requested, etc.
The advantages of this kind of built-in data collection and immediate analysis over Skinner’s teaching machines and Crowder’s scrambled textbooks were abundant and obvious. Whereas PLATO and non-computer-based teaching machines both offered Self-Pacing and Immediate Feedback, a system like PLATO extended Immediate Feedback to the teacher as well, enabling her to quickly see how a student was doing. In 1960, this was a giant leap forward.
During the fall of 1960, the PLATO team managed to create several small lessons to begin experimenting with the system, in elementary number theory, German grammar, and computer programming. Alpert had green-lit the new PLATO project with the idea that Bitzer would initially develop interactive lessons teaching people how to use the ILLIAC itself. Bitzer and Braunfeld did develop some binary arithmetic lessons that were key to understanding ILLIAC programming, but they did not hesitate to venture beyond that subject matter, knowing that PLATO would need acceptance and interest from a broad range of faculty disciplines if the system were to be viable long term.
—
The PLATO team knew that sitting a single student down at a single “Keyset” and “TV Display” connected to the gigantic ILLIAC machine was not going to be a compelling demonstration of the economic viability of computer-based education. The only way CBE was ever going to be economical was if multiple simultaneous students each could sit down at their own terminal and undertake their own lessons at their own pace, regardless of what else was going on on the system at the moment. This meant that PLATO had to become a time-sharing system. By December 1960 Bitzer and company were already working on it, mentioning the concept for the first time in another CSL progress report, describing a plan for “construction of a machine with the same teaching logic…which will be able to handle more than one student concurrently. In this version of the machine, two or more students will be able to work completely independently.” This would become PLATO II.
The PLATO II diagram Credit 11
Bitzer tried to get the University of Illinois to file a patent for their implementation on the ILLIAC of “time-sharing,” but the university botched the patent application. “We filed for a patent application right away,” says Bitzer. “The person who supported PLATO dearly throughout the whole life was the president of the university, David Dodds Henry; he was the president who also managed to get the patent lost in the university. Not because he wanted to, but in those days, universities and patents, you know, it was a curiosity piece. It wasn’t his fault. He wasn’t trying to disturb PLATO—he loved PLATO. It’s just that the university was not set up to handle things like this. You’re a pioneer in many ways.”
A few years later the time-sharing patent application was found, stashed away in some archive. It was re-filed, but it was too late. “As a result,” says Bitzer, “we lost time-sharing. And that’s something we would have had as well.”
Crack open any computer history book or encyclopedia article today and you’ll learn that MIT is credited with the first implementation of time-sharing, and usually there’s no mention of PLATO or Illinois. “We lost a lot of clout,” says Bitzer. It was a tough lesson that he took to heart. He would be ready the next time.
—
Al Avner, an educational researcher at the University of Illinois who would in time join the PLATO project, once described Don Bitzer himself as the original “time-share system.” Those wishing to program on the ILLIAC I, Avner would recall, “would usually be told by Don that whatever needed to be done was trivial and to simply drop by Don’s house that evening to see how. Upon arrival at seven or so, one would find a group with similar interests gathered around the Bitzer dining room table. Don would work on each person’s problem for about five minutes in rotation until three or four in the morning and the ‘trivial’ problem would be solved in a few mind-numbing evenings (everyone was also working normal hours)….Anyone who wanted to use PLATO so much that they were willing to go through the agony of this process was deemed to have a strong enough interest in [computer-based education] that their inputs were of value in designing the system.”
With the arrival of 1961, the PLATO project was gaining speed. The effort to build PLATO II was under way, and Bitzer and Braunfeld were giving more demos all the time. In fact, “the demo” was to become the main thing Bitzer seemed to be doing: showing the system off to whatever interested party would sit down to take a look. In March, the interested party was David Dodds Henry, the president of the university. He put on a conference, called “Improving Our Educational Aims in the Sixties,” at Allerton Park, a secluded retreat among tall pines some thirty miles to the west of the campus. It would mark a milestone in PLATO’s history: the first time the system had been demonstrated remotely, dialing in from a distance.
The PLATO II terminal Credit 12
Bitzer and Braunfeld recruited a young grad student named Rick Blomme to play the role of “student” during the demonstration. “The engineers thought Bitzer was crazy,” Blomme says. The idea was to use a regular phone line and a primitive, hacked-together modem to send the student “keyset” data back to the ILLIAC. But you couldn’t use a phone line for the signal coming to the student’s “TV display”—that would have to be an actual video signal. Cable television did not exist yet, and they certainly were not going to run a fresh coaxial cable from CSL all the way out to Allerton. So they came up with a “kludge,” in keeping with Bitzer’s frugal, pragmatic style. The university ran a television station in town, WILL-TV. It did not have any broadcasting in the morning. But it did that day. Bitzer rigged up the closed-circuit video output from the storage tube, and ran it over to WILL-TV, which broadcast it over its antenna. Had anyone turned on their TV at home that day, they would have seen the same output that President David Dodds Henry and the assembled dignitaries saw on a television screen out at Allerton: the output of a PLATO lesson.
Bitzer gave the main talk about PLATO and then Braunfeld introduced Blomme to the assembled audience as the live, remotely dialed-in demo began. “Here’s a student, now, he’s going to study this little lesson, and push keys.” As Braunfeld lectured the audience about the workings of the PLATO system and the design of the lesson, Blomme would wait to get a signal from Braunfeld to do something. “Now here’s what the student would do, he doesn’t know what to do so he would press the HELP key,” which was the signal for Blomme to press HELP, after which Braunfeld would describe how HELP and AHA worked.
Blomme may have been portrayed at this conference as the innocent student, but in reality he was no slouch at computers. “I started in chemistry and went into engineering physics,” he says, “went to grad school, did mathematics, got a master’s, and that was about as far as I went. They were getting tired of me, and I had gotten involved with computers….My understanding was, in those days there were no computer programmers. People were electrical engineers or chemists and physicists and they programmed the computer to do whatever they needed to do.” The idea that someone would be a programmer—as in, that was what they did, program, full-time—was looked down upon in 1961. Blomme was one of the first to break the mold, and proudly identify as being a programmer.
Blomme would soon be recruited by Larry Stolurow, who had also
presented at the Dodds conference at Allerton, to work on a competing computer-based automatic teaching system at UI. He would be Stolurow’s programmer.
5
Soldering Irons, Not Switchblades
The orchestra played “Arrivederci Roma” as passengers drank, danced, and celebrated the last evening of their westbound transatlantic journey that had begun nine days earlier at a dock in Genoa. The captain had scheduled the ship to pass through the narrows between Brooklyn and Staten Island at sunrise, offering everyone a dramatic view of the Statue of Liberty and the New York skyline as they steamed into the harbor. Twelve-year-old Andy Hanson, his ten-year-old brother, Donnie, his seven-year-old sister, Ardith, and his six-months-pregnant mother, Elizabeth, were looking forward to coming home. They’d packed and gone to bed early so they’d be up and on deck in time to witness the exciting end to their voyage.
Andy’s father, Alfred, a nuclear physicist, wasn’t aboard. After working on the Manhattan Project in Los Alamos, New Mexico, where Andy had been born only months before the first atomic bomb was tested, Alfred packed up the family, moved to the University of Illinois, and became an assistant professor in the physics department and eventually project director for the Betatron, a four-hundred-ton particle accelerator. He began a long, prolific period of research studying the structure of protons and neutrons, research that helped lay the groundwork for the multibillion-dollar, miles-wide particle accelerators and colliders built in more recent times. In 1955, Dr. Hanson was awarded a Fulbright scholarship to help physicists at the University of Turin build their own Betatron. The whole Hanson family would move to Turin for the year. That summer, the family sailed to Europe and began their Italian stay. In just a year, Andy would cultivate a lifelong fluency in Italian and love for Italy. Alfred Hanson was invited to confer with nuclear physicists working on an experiment at the Brookhaven National Laboratory, so he flew back to New York early, and planned to meet his wife and three children when their ship docked in New York Harbor on July 26, 1956.
On the night of July 25 they approached the mainland of the United States, south of Cape Cod. As passengers danced and drank the night away up on the main decks, the four Hansons lay asleep in what Andy describes as “not quite the worst cabins” of tourist class, situated one level above the waterline. At 11:10 p.m., they were jolted awake by a violent crash followed by a chilling, unearthly, grinding groan of metal against metal—EEEEE-zhurrrrrr—a deathly sound that Andy Hanson still recalled more than fifty years later as if it had happened yesterday. He bolted upright, looked out the porthole from his bunk, and saw, from what seemed to be just an arm’s length away, lights from another ship, a moving wall of steel, pass right by the window. That did not make any sense. He immediately slammed shut the cast iron cover of the porthole and tightened the screws. Then he climbed out of bed.
Within seconds, everyone noticed that their ship, the 701-foot-long Andrea Doria, a world-class ocean liner beloved by many celebrities of the time, was listing starboard. Within a minute it was listing 15 degrees. Soon it was listing 20 degrees. Confusion reigned among passengers and crew alike. In the middle of a fog bank, on the final evening of a long cruise from Europe, the Andrea Doria had crashed into something, or something had crashed into it, and now it was sinking.
It turned out the collision was with another ocean liner, the eastbound Stockholm. It had entered the same fog bank that evening, some forty miles south of Nantucket. A series of miscommunications, misunderstandings, and mistaken indicator readings led the crews of both vessels to commit to fateful wrong turns. The Stockholm, sporting a special ice-breaking bow ideal for navigating the ice-filled waterways of Sweden, gouged a massive hole into the Andrea Doria’s steel flank, crushing numerous cabins, instantly killing forty-six passengers, and tearing open several fuel tanks, at the time nearly empty and full of air after the long voyage. The ruptured starboard tanks rapidly filled with ocean water. The combination of empty port tanks full of air and the weight of all that onrushing water into the starboard tanks caused an uncorrectable starboard list that quickly went extreme. Within twelve hours the Andrea Doria would sink to the bottom of the ocean.
The Hansons were about to find out just how low on the totem pole tourist-class passengers were. Steel gates blocked stairways up to the wealthier classes. “It was very well partitioned off,” Andy Hanson says. He compares the class demarcations to scenes in the 1997 film Titanic, where the poorest passengers were kept behind caged-off stairways deep belowdecks. “By the time we got up through our tourist class cabin labyrinth up to the deck, all the lifeboats were gone, there were no rescuers to be seen, so we sat there for a couple hours.”
As they waited for help, their ship continued to list further, slowly drifting away from the Stockholm, which was now invisible in the dark, foggy distance. The crew of the Stockholm would in time realize that despite the fact that the bow of their ship was crushed—there was no longer any bow at all—their ship appeared to be stable and was not going to sink. So they decided to release a couple of lifeboats to rescue Andrea Doria passengers. Out of the dark mist down in the water, Andy made out several lifeboats appearing out of nowhere. Elizabeth Hanson was determined that her children were going to be rescued. “We’re going for the lifeboats,” she declared. The Doria was now listing 40 degrees, the edge of the deck now only twenty or thirty feet above the water. “We went down,” Andy says, “and there were no ropes, no ladders, no nothing, on the deck we were on, no way to get to the lifeboat except jumping in the water.” Andy jumped in first, finding the water warm as bathwater, and swam to the first lifeboat, where he was quickly pulled on board. “My mom says I sat there grinning like a Cheshire Cat,” Andy recalls. His sister jumped in next, followed by a reluctant Donnie, who needed a decisive push from his mother. Then she bunched up her knees to protect her unborn child and jumped in next. They all swam to the lifeboat and got on.
Other passengers followed suit. One small Italian family, frantic to save their four-year-old daughter, Norma, “just went completely crazy,” Andy recalls. The father tossed the child down to the boat before people in the lifeboat were ready to catch her. Andy watched the child fall and hit her head against the side of the lifeboat, instantly knocking her unconscious. Rescuers had to call for a helicopter, which transported the Italians to a Massachusetts hospital. The child died the following day.
The lifeboat made it back to the Stockholm, which, while still afloat, was so heavily damaged that the captain maintained a speed of only two knots. Another ocean liner arrived a few hours later and picked up the rest of the passengers. It would take two more days before the Stockholm would limp into New York Harbor and the Hanson family could be reunited with their father.
More than fifty years later, now Professor Emeritus of Computer Science at Indiana University and former head of the department, Andy still thinks about that night. Over the years he’s figured out the math. For him it all boils down to a difference of 2.7 seconds. That’s all it would have taken. “I had gone back into the records,” he says, “and found that the Andrea Doria was moving at 21.6 knots. And if you do the conversion and you calculate that, that’s thirty-five feet a second. I’m ninety feet away from the collision. Forty-six people died instantly at the collision point. That’s 2.7 seconds. I’m either here or I’m not here. The other ship could have been slower, the Doria could have been faster. Two-point-seven seconds would have been me instead of those other forty-six people.”
—
From an early age Andy Hanson had an interest in science and technology, no doubt influenced not only by his physicist father but also three physicist uncles. In 1955, before his family left for Italy, he’d built a robot of sorts in his garage. “I did lots of things,” he says, “built rockets, blew things up.” When he needed to find or make parts for his latest gadget, it helped to have a father who ran the Betatron, in a building with not only tools galore but a complete machine shop. “We’d get stuck with some kind of gear we
needed, and we’d go over…and machine it at the Betatron.”
Hanson enrolled in Urbana Junior High School and, one night in early October 1957, watched from his backyard as a tiny sunlit speck crossed high up in the sky: Sputnik. “I remember huddling around a little old tube AM radio in our basement listening to the beeping,” he says. Even though he was just in eighth grade, he took a high school trigonometry class over at Urbana High School. In that class he met a kid named Roger Ebert. Susan Gilbert, another student at Urbana High, and editor of the school yearbook, was Ebert’s “archrival,” Hanson says. “[She] succeeded in getting his goat by inserting the aphorism ‘A self-made man who adores his creator’ under Roger’s yearbook picture.” Decades later, Hanson, who regularly attended the Ebertfest film festival in Champaign, bumped into Ebert, who remembered him not only by name, recalls Hanson, but “as the precocious freshman in his junior year geometry class with Miss Bauer, UHS’s legendary geometry teacher.”
Urbana High School participated in a program called JETS, the Junior Engineering Technical Society, a national organization founded in 1950 at Michigan State University with the goal of encouraging high school students to immerse themselves in challenging engineering projects, which, it was hoped, would spark an interest in engineering as a career. Many high schools near Big Ten colleges had affiliations with JETS, and by the late 1950s, thanks in part to Sputnik, some four hundred JETS clubs had been formed. Andy Hanson got involved in JETS at Urbana High, as did several classmates, including Gary Gladding, who would go on to have a long career as a professor of physics at UI.
Don Bitzer wound up being one of the UI faculty sponsors for the Urbana High School JETS program. Just a year into the PLATO project, he was already sponsoring a few kids at University High School (known to everyone as “Uni High” or just “Uni”), located on the other side of Springfield Avenue from CSL. Bitzer had already tasked two Uni students, George Frampton and Steve Singer, to help him with PLATO. According to Hanson, Bitzer was having trouble raising money for the nascent PLATO project, and also had trouble attracting graduate students or postdocs to pitch in and help. “Frampton and Singer were being trained,” Hanson says, “as programming assistants for Bitzer to do projects he was unable to get done by other people. I remember him mentioning this to me many times, ‘I don’t even know why I would bother with college students and postdocs, I’ve got the smartest high school students in the universe right here, I’m just going to grab ’em!’ ”
