In a bit of eerie foreshadowing, this episode of The Computer Chronicles from January 1984 opened with a discussion of a global pandemic. Not a real one, of course, but a computer simulation. Stewart Cheifet and Gary Kildall played with a game called Epidemic, released by Strategic Simulations. Cheifet explained this was an example of how someone could use a computer to choose among alternatives, make a decision, and establish a complex strategy for solving a problem. In this case, the problem was a global pandemic caused by a virus brought from outer space by meteors. Epidemic provided the player with information about the spread of the virus and offered alternatives, such as developing a vaccine, for containing the pandemic.
The actual topic of the episode was simulation games, but rather computer-aided decision making in business. Also known as decision support systems. Kildall said this was really a catch-all phrase for a lot of different concepts. But the basic premise was that the computer was helping a manager or decision-maker with long-term strategies and providing an “able assistant.” The underlying technology could involve the use of expert systems, relational databases, or even spreadsheets.
Cheifet then provided his customary narration over some B-roll footage. He reiterated that decision support systems encompassed a broad range of analytical tools, from forecasting programs that calculated trends and predicted outcomes to expert systems that could draw its own conclusions and make decisions. Whether or not these decision aides required some form of artificial intelligence, they were already in routine use, having proven their efficiency and accuracy in a number of applications.
For example, Cheifet said a “major American railroad” currently used an in-house decision support system to assist in long-term financial and operations planning. Based on forecasts of future demand, the system reviewed records of train capacities, labor costs, taxes, fuel costs, and revenues to predict the financial impact of alternative strategies.
Cheifet pointed to another early decision support system called Prospector, which analyzed the mineral content of geological formations to recommend drilling sites. Prospector had many features associated with artificial intelligence. That is to say, it took user inputs regarding various rock types and used that information to draw conclusions and make independent recommendations. However, Cheifet noted that while Prospector relied on “probabilities” to draw its conclusions, the more common types of decision support systems in use today required “known facts.” But future research was likely to narrow the gap between programs that laid the groundwork for decisions and those that could make them.
Expert Systems vs. A Well-Trained Assistant
Back in the studio, Mike Thoma and Steve Weyl joined Cheifet and Kildall. Weyl was a co-founder and vice president at Syntelligence. Kildall asked him to provide a commercial example of a decision support system. Weyl said his company specialized in what were known as “knowledgebase” or “expert systems,” which tried to bring the actual judgment of an expert to the desktop of a manager or business professional who needed to make a particular decision. For example, Syntellience was now working on a system to help commercial lending officers at banks decide whether to make loans to small companies.
Cheifet then asked Thoma, a manager with Management Decision Systems, if a decision support system had to be an “expert” one like what Weyl described. Thoma replied no. He joked that when he asked his 78-year-old father, who worked as a consultant, if he needed a decision support system, his father replied no, he made the decisions. What the father wanted was software that could provide an “alternative evaluation” of a situation. Thoma said this boiled down to providing a “very well trained assistant” who would carry out an analysis at the user’s direction.
Cheifet asked about the importance of interactivity to a decision support system–did it have to respond uniquely to the questioner? Thoma said he believed that interactivity was important. The decision-making process was one of innovation, he said, meaning that you had an idea and needed feedback. The system’s responsiveness, in turn, would help generate additional ideas.
Kildall noted these decision support systems required a connection to a larger computer system with database access to provide the basic facts and figures. Thoma agreed that some sort of computer connection was required, although the size depended on the problem. Cheifet asked Weyl if Syntelligence’s system depended on a mainframe. Weyl said yes, the systems they had initially developed required a mainframe, but given the present direction of personal computers, he expected that within a few years “at most” their systems would run on future PCs.
With respect to interactivity, Weyl added that Syntelligence had spent a lot of time ensuring that its systems provided full explanations of how they reached their conclusions. He said managers could not simply look at the computer and say it “told me to make that loan.” The computer had to work like a staff assistant, and like a person it needed to give a full explanation of its reasoning.
Cheifet then asked Thoma to demonstrate his company’s product, Express EasyScan. Thoma created a hypothetical scenario where the user was looking to buy a software company. The software initially identified 62 possible companies from a database provided by Standard & Poors. Initially, the 62 companies are listed in alphabetical order by their stock ticker. Thoma then entered parameters to narrow down the field: Specifically, he queried for just those software companies that had reported a sales growth rate of 30 percent over the five-year period up to 1982. This brought the list down from 62 to 12. Thoma then ranked those 12 using an additional criteria, profit margin. Cheifet then asked for a bar graph comparing just the top-two companies by profit, which Thomas produced.
Taking a Personality Quiz to Improve Your Sales Performance
Jim Chapman joined Cheifet, Kildall, and Weyl for the last segment. Chapman was a vice president at Human Edge Software. His company produced The Sales Edge, which unlike Express EasyScan was designed to work without a mainframe on a standard IBM Personal Computer. Chapman explained the software provided a step-by-step strategy report for sales, negotiation, and management.
The salesperson would start by answering 81 pre-loaded questions about themselves, essentially generating a snapshot of their characteristics and attitudes towards sales. Kildall semi-seriously answered a few of the questions as a demonstration. The questions were actually statements that the salesperson would either “agree” or “disagree” with, such as:
- Sometimes a sales call makes me nervous.
- I like the freedom to travel that a sales job gives.
- I enjoy swapping stories with customers.
- I am sensitive to criticism.
- I enjoy putting one over on a customer.
- I like talking in company meetings.
Chapman explained the final answers were saved to a diskette. He then showed how the salesperson could call up a previously completed customer profile, in this case for a hypothetical customer named Monroe. Monroe’s profile displayed a number of adjectives describing him, such as that he was “not talkative” and “apprehensive.” Chapman said this information was then used to generate a more comprehensive report about the customer. Cheifet clarified that this was used to determine what strategy the salesperson should take in dealing with the customer. Chapman said that was exactly right. In particular, when a salesperson met the customer for a second time, they would already have a feel for what they were like, what their strengths and weaknesses were, and so forth.
Kildall asked how successful the software was in an actual sales situation. Chapman claimed it was very successful. He said he used The Sales Edge himself and so had approximately 1,000 customers around the country during “beta testing.” And the feedback was that the program was useful.
Cheifet pushed back a bit, asking if this approach was not “superficial,” and if you really could make a business decision based on a bunch of quick “yes or no” questions. Chapman insisted you could. He noted this was actually a cross-fertilization of two areas of research: personality and sales. He said the software helped a salesperson select the most appropriate strategy based on how a customer thinks.
Kildall reiterated the importance of this type of software telling the user why it made a certain recommendation. He said it would be hard for him to trust a program on a disk that told him what way his business should go. Chapman pointed out that The Sales Edge actually required multiple diskettes. For example, a separate disk was necessary to generate the final customer report, which Chapman demonstrated.
Cheifet asked Weyl for this thoughts on how far he thought decision support systems could go. Would we see programs like this used in making legal or governmental decisions? Weyl said he thought that was still 3 to 5 years away. The types of events that would influence the head of a government were quite a bit more difficult to model. Kildall noted there was a danger that people could become too reliant on decision support systems in the future. Weyl pointed out that was a danger with computer systems in general. Ultimately, he believed these systems would help improve decision-making, but they couldn’t be followed slavishly.
Commodore Profitable Despite Losing Executives
This edition of “Random Access,” recorded in late January 1984, was the first one reviewed here not hosted by Stewart Cheifet. Susan Bimba, another KCSM-TV employee, served as the guest presenter.
- The federal government was investigating 12 semiconductor manufacturers suspected of selling faulty and inadequately tested chips. The focus of the investigation was National Semiconductor, whose officials believed criminal and civil charges were imminent. Bimba noted that in 1982, National Semi admitted its chips did not receive the “required amount of heat-stress testing” required by its government contracts.
- The California legislature was considering an anti-hacker bill that would make it a misdemeanor to break into any computer system even if there was no malicious intent; if there was such intent, the charge would be bumped to a felony.
- Commodore reported a second-quarter profit of $50.1 million, more than doubling its figures from the previous year. Meanwhile, four Commodore executives resigned from the company just two weeks after CEO Jack Tramiel left for Atari, although Commodore claimed there was no connection between the two.
- Commodore also announced it would delay the release of its 264 computer, which was originally scheduled for April 1984.
- Following last week’s introduction of the Macintosh, Apple President John Sculley reorganized the company, reducing the number of divisions from 6 to 3. The new divisions were (1) Apple II & Apple III computers; (2) 32-bit systems, i.e., the Macintosh and Lisa; and (3) accessory products, including printers, keyboards, and modems.
- Bimba said Apple also planned to open a new factory in Mexico to comply with that country’s requirements that computer sellers operate local manufacturing plants. Apple already accounted for 40 percent of Mexico’s personal computer market.
- Hewlett-Packard and Vitalink Communications announced plans to start a hi-speed satellite communications network that would allow personal computer users to access a network without a telephone.
- Storage Technology Corporation laid off 400 workers at its Santa Clara plant, giving employees just 2-hours notice. Bimba said the plant had been working on a large, IBM-compatible mainframe, but a decline in sales during 1983 led to the layoffs.
- Finally, Bimba said, if you “joined the chorus that’s saying video games are a thing of the past, you may have to change your tune soon.” Bimba noted computer games were now becoming more popular and that a number of Silicon Valley firms had switched from video games to computer games to take advantage of this new trend. That said, gaming enthusiasts said both could coexist.
“Human Edge” Not Enough to Become Profitable
The most interesting thing to me about the two products demonstrated–Express EasyScan and Sales Edge–was the wide disparity in price. And I’m really underselling that point. As noted in the episode, Express EasyScan was mainframe software. And it carried a mainframe-level pricetag: $50,000. That was according to a February 1985 article in ComputerWorld, which said the financial consolidation and reporting system module–i.e., the S&P database of company information–added another $120,000.
In contrast, The Sales Edge was desktop software that you probably could have bought in any computer store. An April 1985 report in the Chicago Tribune, said The Sales Edge only cost $250. And it worked on the IBM PC and Apple computers, including the Apple IIe and the Macintosh. (The Apple IIe title screen is what you see in the thumbnail for this post.)
Unfortunately, the Tribune article also noted that Human Edge Software itself was “awash in red ink,” having posted a $1 million loss in 1984, although company founder James Johnson said he expected to be profitable in 1985. Johnson himself was a clinical psychologist by training, and he previously “developed computerized diagnostic systems used by psychologists and psychiatrists,” before starting Human Edge. The Sales Edge was Johnson’s first commercial product, which was followed by Management Edge, Communications Edge, Negotiation Edge, and Mind Prober, according to the Tribune.
My guess is that Human Edge never did attain profitability. It went out of business in 1988.
Notes from the Random Access File
- This episode was recorded on January 24, 1984, and is available at the Internet Archive.
- Management Data Systems (MDS) only lasted until around 1986, when it merged into another company, Information Resources, Inc.
- Mike Thoma stayed with MDS until 1985 when he moved over to Oracle Corporation as its vice president of marketing for three years. He went on to hold a number of marketing positions in the tech industry, including eight years at Actuate Corporation from 1999 to 2007. In recent years, he’s moved into executive coaching and career mentoring.
- Steve Weyl was a co-founder of Syntelligence and left the company after it went bankrupt in 1991. He worked at Apple for a couple years after that and later served as President of Ricoh Innovations from 1996 to 2007. He’s currently a principal at Xeolux, LLC, a brand and intellectual property development company.
- From what limited information I could find, “Random Access” guest presenter Susan Bimba passed away in 2017.
- The Commodore 264 project mentioned by Bimba was eventually released as the Commodore Plus/4 in June 1984.
- National Semiconductor was indeed indicted by the U.S. Department of Justice in March 1984. The company agreed to resolve both criminal and civil charges in exchange for paying a $1.75 million fine.
- Looking at the instruction manual for Strategic Simulations' Epidemic, I learned that apparently you can kill an airborne virus by setting the atmosphere on fire: