On the use of computers for simulation in Stafford Beer's projects, including a witnessed encounter between Beer and Warren McCulloch.
John Ludley, a former computer simulation expert and colleague of Stafford's at United Steel and (briefly) SIGMA, shared his experiences in an interview on August 10, 2021. The conversation revolved around the use of computers for simulation, interactions between Stafford Beer and Warren McCulloch, as well as Stafford's reluctance to engage with the computer science community following his experience in Chile. The interview has been edited for clarity and conciseness.
Evgeny: John, could you tell us a bit about your background and how you first got involved with computers and engineering?
John: My name is John Ludley. I attended Bradford Grammar School, where I was good at maths, physics, and chemistry. I chose to study civil engineering at Leeds University based on these interests and stayed there from 1953 to 1959. During that time, the university installed its first computer. This was during the early days of computers being introduced to colleges, and I stayed on to do research in stress analysis. I took advantage of the computer and used it for my research. In fact, I wrote a couple of programs for calculating stress analysis in slabs with beams, which turned out to be interesting for engineers and particularly for United Steel .
A friend of mine who had been a graduate apprentice at United Steel and had met Stafford Beer was very impressed by his work. I went to Brown University in America for a year to study Operational Research . My friend Stuart told me about United Steel, so I contacted them while I was in the US. Stafford happened to be visiting for a meeting with Warren McCulloch at MIT . I met him, and after having lunch and listening to him talk, I was offered a job. I joined United Steel when I returned to England in 1960, and that's when it all started.
Evgeny: Can you describe your first meeting with Beer and what kind of impression he made on you?
John: I honestly can't remember much about Stafford's appearance at the time, though I know he didn't have a beard back then. He was very personable. During our first meeting, he did most of the talking while I listened. We met Warren McCulloch in an empty MIT classroom, and I listened to their conversation, though I didn't understand much of it at the time. Afterward, we went for lunch, and our initial meeting lasted about three to four hours. I didn't see Stafford again until I moved back to England and joined United Steel.
Evgeny: What kind of projects were you involved in at United Steel, and what methods did you employ to improve efficiency?
John: At United Steel, we focused on operational research and cybernetics. Although there wasn't much cybernetics happening at the time, we used operational research to examine different projects and improve their efficiency through simulation. We employed linear programming and integer programming for various calculations, which can now be done using Excel. I also developed some simulation software and worked on critical path analysis, a linear programming technique, for which I wrote additional software. My main work involved writing software for stress analysis in steel structures for the people in Scunthorpe. It was a significant project at the time, though I wasn't sure if it was revolutionary. I also used linear programming for different types of work within Operations Research, separate from the main project I had.
Despite the skepticism from senior management, we did a lot of good work. Stafford, in particular, had very innovative ideas about cybernetics that I didn't fully understand. He attempted to get United Steel involved in more advanced projects, which are now referred to as AI. Back then, he was exploring concepts like using a tank full of microorganisms called Euglena to create a management style for a company. These ideas eventually evolved into the AI techniques we see today, such as facial recognition.
John: Tocher was undoubtedly very talented. I worked more closely with him than with Stafford and enjoyed collaborating with him. However, he didn't get too involved in Stafford's projects, which he considered highfalutin. Tocher focused on simulation and had people write simulation programs for him. As for myself, I wrote some simulation programs, but not using the GSP . Instead, I used Pegasus code and developed a linear programming system.
Evgeny: Did Stafford's eccentric side come out strongly during your time at United Steel?
John: Absolutely, Stafford had quite eclectic interests, particularly in the field of cybernetics, an area that he believed could be put to great use at United Steel. His fascination extended to projects that today would fall under the umbrella of artificial intelligence. One vivid example of this was his attempt to emulate a cooperative environment within a company by studying a tank full of microorganisms he referred to as Euglena. He imagined the patterns of their interaction could inform a new management style for businesses.
The idea was incredibly futuristic, almost baffling for that era, but you can see how it has morphed into today's practice of using vast amounts of data in AI applications, such as facial recognition. Even though the techniques of today far exceed what we had at our disposal back then, I believe Stafford's pioneering ideas were instrumental in shaping this trajectory.
One thing that remained clear was Stafford's enthusiasm was entirely genuine; it wasn't some spectacle, but a manifestation of his inventive mind. While Tocher, who I consider to be the leading scientific brain at United Steel, did not involve himself as much in Stafford's experimental work, he held a significant role in simulation, another cutting-edge area at the time. But regardless of the different paths they pursued, both were instrumental figures in United Steel's pursuit of innovation.
John: By the time I joined, I had a PhD in engineering. Most people at Sigma had degrees in some other scientific or technical field, or even as classicists. I simply wrote Stafford a letter one day saying: "I'm bored. Can I come and see you?" Two lines, like that. I've still got his reply. I've still got his letter, which said: "Dear John, no surprise. Yes, you can. Stafford", so I went down and joined in there. Jonathan Rosenhead followed me, and Mary Scott, who was Stafford's PA at Sheffield, also came and worked for him at Sigma. In many respects, she knew Stafford closer than the rest of us, because she was in the room with him or in his office.
Evgeny: What kind of work were you involved in at Sigma?
John: I was at Sigma for 18 months. I was writing a computer program, a linear programming system for Elliott automation to put on their computer there. That's a project I had working for Sigma. Writing software. Sigma was related to a Parisian group. They acted almost as an offshoot of that, and in Paris, they had a CDC 3600. For more complex calculations, we had to use a bigger machine, and sometimes we had to go to Paris. I don't think I went to Paris, but one or two of my colleagues would go to Paris to use the CDC Machine for doing the calculations there.
I remember running a linear programming project to find out whether it was cheaper for them to fly to Paris to do it or to rent the machine in London because Cycon had a CDC Machine. That was just a linear program, a sort of fun thing to do. When Sigma thought about buying a proper machine, I felt I needed to be somewhere where they had a bigger computer, because that was my real interest. Then it so happened that I found out that the National Engineering Laboratory in Scotland, a government department engineering laboratory, had put in a Univac 1108 machine, and they were looking for somebody to run it. I applied for the job, and because I was an engineer who knew about computing, I got a fairly senior government position at a very young age. I was about 31.
Evgeny: How was your communication with Stafford Beer in Sigma and United Steel? Did you notice any differences?
John: Maybe it was slightly closer in Sigma because the senior management from the United States didn't take up as much of his time. However, overall, I don't think there was a significant difference. In Sigma, I talked more with David Owen, and in United Steel operational research, I talked more with other colleagues.
Evgeny: Can you share some memorable moments or experiences you had with Beer?
John: One funny memory I have is from the first time I met him in Boston. It was a rainy day, and I had an umbrella that automatically opened when I pressed a button. This impressed Stafford, and he later mentioned the encounter in his book, although not by name. Most of the time when I was with him, I would listen to him talk with Warren McCulloch, and we would have lunch together.
Evgeny: Can you describe the qualities that made Beer stand out as a thinker, even though he wasn't a traditional scientist or mathematician?
John: He was a man of exotic great ideas. He surrounded himself with good scientists like Tocher who could do the work, but his ideas were more in the mind, forward-looking. It's difficult to put your finger on what he was. He wasn't a great scientist. He was a good thinker, a great one even.
John: Yes, I did. I visited Stafford at his cottage in Wales and asked if he would come and talk to the British Computer Society. He declined, saying that his ideas might be a bit old-fashioned for the London audience, even though he was a professor in Canada at the time. We had a nice chat before saying goodbye. He wondered if his ideas might be picked up by the treasury or the statistical office where I was working. However, I told him that it was unlikely, as I was just a computer guy there.
Evgeny: What did you make of him choosing to live in that cottage of all places?
John: Well, he was a professor at two universities at the time, and it seemed like he was living there temporarily. I never asked him the exact reason, but one possibility is that one of his daughters was close by, and the friends who were visiting had a connection to his PA, Mary. She was his secretary and helped to look after him. So, perhaps he needed somewhere to stay and be looked after, as he was spending a lot of time abroad, mostly in Canada.
Evgeny: Were there any political discussions at Sigma and United Steel?
John: Politics was never really mentioned. I mean, some departments might have been influenced by the political views of senior people, but to my knowledge, that wasn't the case.
Evgeny: How did your involvement with Beer impact your overall career trajectory?
John: The fact that he offered me the job in Sheffield led me to meet a lot of interesting people. Then I went to Sigma, met more interesting people, and ultimately got a job. The background I developed in computing and engineering enabled me to get a job with the government at NEL. I was there for 10 years and then moved to be the head of computing in the statistical office for the last 20 years. So, in a sense, it created the route for me to climb that career ladder.
A steelmaking, engineering, coal mining group based in England. Nationalized and succeeded by the British Steel Corporation. More
A problem-solving and decision-making discipline using advanced analytical methods to help manage an effective organization. Originated in Britain during WWII for integrating technologies (such as the radar) into warfare tactics. More
Warren McCulloch: An American neurophysiologist and cybernetician, he contributed to the founding of neural networks and computational neuroscience. More
MIT: Known for its scientific and technological contributions. Criticized during the Vietnam War for its ties to the military-industrial complex, sparking debates on research ethics and social responsibility. More
A British computer scientist, who wontributed significantly to the birth of simulation and event-based analysis. Working under Stafford Beer at United Steel, he developed groundbreaking simulation methods that substantially advanced industrial operational processes, embedding him in the history of system sciences. More
General Simulation Program (GSP) is a pioneering discrete-event simulation framework developed by KD Tocher and his team circa 50 years ago. Initially envisioned as a tool for modeling diverse steel plant operations, it evolved into a versatile, high-level simulation program widely used in operational research.
Science in General Management (SIGMA): an Operational Research consulting firm ran by Stafford Beer in the early 1960s after he left the steel industry.
A British operational researcher recognized for his work on problem structuring methods. A friend and colleague of Stafford Beer, Rosenhead influenced the practical application of operational research in complex decision-making scenarios. More
A professional body representing the UK computing field, promoting computing knowledge, setting industry standards, providing certifications, and fostering networking. More