Title: Artificial Artificial Intelligence
Date Occurred: Spring 1983
Date Written: January 6, 2006
Written By: Joel T. Kant
Copyright (c) July 19, 2019
It was nice using the modem that John Stilwell lent me during the spring semester of 1983. He had lent it although he was attending University of Wisconsin-Madison while I was attending University of Wisconsin-Platteville. He had replaced his 300 baud modem with a 1200 baud modem to communicate much faster, four times faster, with his Commodore Vic-20 through the telephone with the various mainframe computers at his university. He might have been able to sell his old modem, but not for much. The advantage of speed at 1200 baud was so great that the selling price of 300 baud modems had plummeted.
I was in a financial tight spot as a result of having bought the Vic-20 with nearly all my savings about nine months earlier, then having had no job the previous summer to get more money. The unexpected one-time small academic scholarship and my part-time math tutoring job had barely kept me afloat, but at that time I could not afford even a 300 baud modem on my own. So far, I had not had to sell my own Commodore Vic-20, but it was a near thing and would continue to be so. I was glad I still had it because it worked wonderfully well for some of my classes. Wordprocessed lab reports rather than handwritten seemed to improve the grades I got on them.
With my Vic-20 and the borrowed modem, I communicated through the telephone with UW-Platteville's PDP mainframe. It had e-mail, but only to other users of that mainframe. Dr. Paul E. Gray of the electrical engineering department at UW-Platteville liked sending his students e-mail. Not all his students regularly checked e-mail in 1983. He explained that he was trying to instill in us the habit of regularly checking it. He certainly was years ahead of the curve on the importance of that!
I checked my e-mail more regularly than most students because I could stay in my dorm room to do it using the modem and my Vic-20, while most students had to physically walk to the computer center to use a terminal.
John Stilwell's former roommate Alan Raichel still lived in their former dorm room in Wilgus Hall. Alan had had a new roommate the last semester, but that student had not returned. Therefore, he had yet another new roommate this semester. Neither roommate enters into this story, nor does my own.
The previous semester, Alan had at great expense purchased a Commodore 64 computer within a month or two after it had been officially introduced. Initially, he had been very proud of it, telling me many ways it was vastly superior to a Vic-20. Unfortunately, it had been broken within weeks. My working Vic-20 was much better than his broken Commodore 64! His repairs had been covered under warranty, but Alan had little use of it all last semester. Just four or so months after he got it, the retail price of the Commodore 64 had dropped from six hundred dollars to under two hundred and fifty dollars! That was an astonishing drop to less than half the original cost in just one semester, foreshadowing how computer prices would plummet in the future. Alan said that he would not have been so upset if he had gotten much use out of his Commodore 64 before that had happened. Sadly, he had not.
During the Spring semester, though, his Commodore 64 system was finally up and running well. Apparently, it really had been repaired rather than replaced. He had bought something with his system that I was jealous of, which was a disk drive. I had something he did not, though, a printer.
Alan's method of getting printouts was the same as John used to do the previous year. It was to upload the files through the modem to the PDP mainframe, then print them from there. It worked, but I vastly preferred having my own printer. Most of the paper loaded in the printers in the computer center had green bars on it. On my printer, I printed out on crisp, clean, white, albeit still tractor-feed paper. There were perforations to tear off the tractor feed holes on my paper. The paper the computer center used did not have perforations. Alan used a paper-cutting guillotine to trim them off if he wanted them off.
Alan liked to point out that for the price of my printer, I could instead have gotten a disk drive, which was vastly superior for storing data than a tape recorder. He was essentially correct (although hard to really work out because I bought my system used), but I loved having a printer even if he had a workaround solution for getting printouts without his own printer. On the other hand, he loved having a disk drive, even if he complained bitterly about the terrible design with a slow communicating serial bus.
For accessing many files, a disk drive is much better than tape! (Consider in 2006 if I want to find a scene halfway into a movie on VHS tape. It takes much faster fowarding and searching. On a DVD, there is a scene index and I can jump right to a scene without fast-forwarding. Using many computer files on cassette tapes is similarly more difficult than using them on a disk drive.)
Alan's 300 baud modem was of a type seldom seen today. It had rubber cups that the phone receiver was pushed down into. The receiver had to have the correct shape for it to work, but fortunately the dorm room telephones had the correct shape. For John's old modem that I was using, I ran wires right into the phone itself, probably violating some dormitory rule doing it.
One day, Alan Raichel, Dan Main, and I were sitting around talking about somebody I will not name.
Alan said, "I was told that he owes hundreds in phone bills."
I asked, "Hundreds? What was he doing, calling other countries?"
Alan answered, "No, calling computer bulletin boards."
I remarked, "I read something about those in a magazine, but I didn't really understand it."
Dan and Alan both looked perplexed by my ignorance.
Alan explained, "You know how we can use the modem to call up the PDP. E-mail can be sent, files exchanged, and things like that. Computer bulletin board systems, or BBS's for short, are like that. They are popular in Madison for trading software."
I asked, "Do they cost money to use?"
Alan shrugged, "Some do and some don't."
Dan put in, "None are local calls from Platteville. The bill for the long distance phone calls would kill you."
"I tried a free BBS when I was home over break. I found it easy to lose track of time, which could lead to huge phone bills if it were long distance," Alan said.
I knew Alan's parents lived in a suburb of Madison, so he could presumably try a BBS without it being a long distance phone call when there.
Alan and Dan told me some more about BBS's.
Alan then said, "If I had an autoanswer modem, I have software from a user group that would let me set up a BBS in my dorm room."
I was surprised, "Really? John's new modem does autoanswer and autodial."
For the modems that both Alan and I had back then, one had to dial the phone manually. The computer would not dial for you. Newer modems, like John's new modem, let the computer dial itself. That was what was meant by autodial. Autoanswer meant the computer could answer the phone similar to how an answering machine does it.
Alan replied, "I think his will do autoanswer. Even if I had a modem like that, I would not really want to set up a BBS, though. I want to keep my phone for use as a regular telephone."
Dan suggested, "I read some people have BBS's only open at certain times, like nighttime after they go to bed."
Alan said, "From what I heard, people who try that get modem calls all day long. Notices about times of operation are not obeyed. I find it interesting that I could set up a BBS in my dorm room, but I won't."
Dan remarked, "One floppy disk on your drive won't hold much anyway, Alan. What you'd really need for a BBS is either several disk drives or a hard disk."
Alan admitted this was so.
I asked, "Can you even get a hard disk for a Commodore 64?"
Alan described a way to do it. I think it involved buying an interface called an IEEE-488 bus, then buying a hard disk that worked with that. Alan's finances were not nearly as bad as mine, but this was just idle speculation. It would have cost a lot of money. I had only read of hard disk drives for personal computers in magazines at that time, not used or even seen them. UW-Platteville's two mainframes obviously had hard disks, though. We all certainly knew about hard disks.
(By the way, in the late Eighties, a number of companies sold hard disk drives for the Commodore 64, but most used an interface called SCSI rather than the interface Alan talked about years earlier.)
Alan then mentioned, "Even without BBS software, I could communicate between my Commodore 64 and Joel's Vic-20. The terminal programs we already use when working with the PDP would do that."
That gave me an idea.
Dan used personal computers, but did not own one, so I directed a question to Alan, "Do people ask you to make your Commodore 64 work with artificial intelligence?"
Alan chuckled at that. Artificial intelligence or AI for short was popular to write about in computer magazines and science fiction stories, but so far it had not amounted to much. Even what did happen with AI took a lot more computing power than a lowly Commodore 64!
Alan admitted, "Occasionally, but only by people who don't know anything about computers."
I said, "A few students have asked me to do AI on my Vic-20."
Dan and Alan had a good laugh at that. The idea of AI on a Commodore 64 was silly enough, but on its predecessor machine, the Vic-20, was absurd.
I told them, "I show them Eliza."
Alan and Dan both knew what that was without any explanation.
Dan asked, "Do any fall for it?"
I said, "Some do for a little while. So far, everybody has caught on within about ten minutes."
There are many versions of Eliza. I had gotten one from Alan. Alan in turn had gotten it from a PET Computer Users Group Disk. The Commodore PET computer was the predecessor to the Commodore Vic-20, although was in some ways superior but also much more expensive. Eliza was a simple BASIC language program without any fancy sound or graphics. With only minor modifications, any computer that ran BASIC and had sufficient memory could run Eliza. (According to Big Computer Games, Edited by David H. Ahl, Published by Creative Press, Inc., (c) 1984, pp. 20-24, the first version of Eliza was written by Joel Weizenbaum in LISP at an unspecified date, then the first version in BASIC was written by Jeff Shrager in 1973. The book says that a version originally appeared in the July/August 1977 issue of Creative Computing. The included code states it is edited by Bob Anderson. Therefore, it is not surprising that I remember many versions of Eliza being around for many different computer systems by 1983.)
Eliza was a program that made the computer pretend it was a psychiatrist.
The start of a typical session would go something like this...
Computer: Please tell me your problem.
User: I hate my dog.
Computer: Why do you hate your dog?
User: He bites me.
Computer: Does it bother you when he bites you?
What the program was doing was far less sophisticated than some of the dialogs would have many people expect. It could recognize a few words like "my." For simple sentences, it knew where the subject, verb, and object should be without any idea of their actual meanings.
The previous conversation might instead have used nonsense words like this...
Computer: Please tell me your problem.
User: I ggrtl my snlzp.
Computer: Why do you ggrtl your snlzp?
User: He qwerto me.
Computer: Does it bother you when he qwerto you?
Eliza had no spellcheck. While it did search for words like "my" to change it to "your," most other words it only parroted right back. The seeming artificial intelligence was just a clever trick.
I found it amusing how long it could take some people to catch on to what the program was really doing.
I then told Alan and Dan a story about a play that I had seen back in Junior High School.
On the stage was a box that could hold a refrigerator or other large appliance. It had aluminum foil glued all over it. It had what looked like blinking Christmas lights on the side. It also had a large sign on it labeling it, "Computer." It also had a large slot like a mailbox. It was labeled, "Input." It had another slot or door labeled, "Output." Inside the box was a speaker and tape player. Sound effects were played as though it were a machine shop mixed with various electronic beeps and hoots. Of course, it was obvious there was a man inside. This box fooled nobody, but it was not really supposed to. The setup was done for a comedy skit. A supposed programmer or user would write a question about a teacher, coach, or student. He would put it in the "Input" slot. The volume on the sound effects would get louder, then a paper would pop out. He would then take that and read it. It was a bad joke about the teacher, coach, or student that the question had been about.
Alan and Dan understood what I was getting at.
I left them in Wilgus Hall and went back to my own dorm room in Morrow Hall. I got the Vic-20 computer and modem ready. I then called Alan's room.
Alan answered, then put the receiver into the rubber cups of his modem back in his own dorm room.
Back in my room, characters appeared on my screen as if by magic. It was Dan typing on Alan's computer, even though Alan set up the connection. I typed on my machine. Text appeared from Dan about the communication working fine. Alan had been correct about how well this would work. We already had the same parity and so forth set because of communicating with the PDP, so this took almost no effort.
My roommate was not around, but this would not have worked on him anyway. He had seen me use the modem to the PDP many times.
I then wandered the hall looking for a victim. I found one student sitting in his room. I told him about the AI program.
We went back to the room, where he tried it.
Dan started with a generic question.
The student typed in something.
The response appeared on the screen.
The student was not fooled, saying, "You've got somebody else typing this somehow. I don't see anybody else around. How on Earth are you doing this?"
He got off the chair to check the room next door.
After he got back from his check where he found no computer in the room next door, I conceded that this was a trick. I explained about the modem.
The student was not upset, but amused. He wondered if others might be fooled. We found somebody else to bring in to try it.
It went much better with the second victim, much to the amusement of myself and the first victim.
The second person was typing questions and responses for over five minutes without apparently realizing there was another person doing the typing!
(A joke like this would hardly work today in 2006! There is hardly anybody around who has not heard of a modem letting a computer communicate over the phone line by now! It was different in 1983, though. Some knew, but many others did not.)
I was delighted how things were going, fighting back laughter.
In the next response on the screen, though, the font on my computer changed color!
A question appeared, "Did your font change color?"
The second victim jerked back and asked me, "What? Why would the computer do or ask that?"
The gig was up!
What Dan had realized is that the Commodore 64 and Commodore Vic-20 could make letters in eight or sixteen different colors. One simple special character would switch the color. He had seen that done and done it himself in BASIC programs. It had occurred to him that the same technique should change the font color on my computer when he was communicating to it by modem.
He was right. That is just what happened. Well, almost what happened. I must admit the font only shifted to a different shade of gray because I used a black-and-white television as a monitor.
Given how well the joke was going with the second victim, I wish Dan had picked another time to try it.
The two victims then spread word around the dorm about the joke. After that, nearly everybody there seemed to know what a modem did, so I could not do this trick again.
Back to Joel Kant Home Page: Joel Kant Home Page.