He was out of discussions with people in industry that I began to look at a different class of problem. In

the early 1950s, Jay had been working as a consultant to General Electric in its engineering department when its management team approached him to help solve a different problem related to its boom and bust performance. Jay used his expertise in building feedback control systems to create a map of all the elements in General Electric's business such as inventories, workers, overheads and so on, and showed how they were all dynamically connected over time, so that the company could spot pressure points and prevent them becoming problems. Working in MIT's cutting-edge mainframe computer labs, he then translates his sketches into computer code, creating a new programming language to help simulate and test different management policies. He called this new branch of science industrial dynamics. Later, a chance encounter with the former mayor of Boston led him to apply this new approach to cities. In

the late 1960s, John Collins came to MIT and by chance took an office next to mine. I said to him, wouldn't it be interesting if we would combine the background that we've had in the corporation with the knowledge of people like yourself in the cities and see if we could come to any better understanding.

Boston, like many cities in the 1960s, was suffering from a rise of poverty and racial inequality. Just as he had done for US corporations, Jay created a map of all the important elements of the city of Boston and how they influenced the well-being of its citizens. As in physics, Jay liked to say that every action provokes a reaction elsewhere in the system. Or to use the engineering term, feedback. Often, policymakers take an action they think will solve a problem, say, building more roads to alleviate traffic, but in so doing they make the problem worse by inducing more cars into the city. His resulting book, Urban Dynamics, came to be used by city planners around the world and would later inspire the computer game SimCity. Fellow MIT professor Carol Wilson, who happened to be a member of the Club of Rome, thought Jay's pioneering work on modelling complex systems sounded like it could help solve Aurelio's problematique. So he invited Jay to accompany him to Bern that summer. And so on that warm June evening in Switzerland, as Jay listened to the increasingly fraught discussions among the Club of Rome members, he took his knowledge of modeling the dynamics of industries and cities and sketched out the first ever world model right there on a set of Swiss napkins. And

I told them they could come to MIT and learn more about this, but they would have to come for two weeks or not at all, because I knew that it would take two weeks for them to really understand. And they agreed. They agreed there at midnight that evening that they would come. They would come three weeks from that day. Jay

touched down in Boston at almost the exact same time as Dana and Dennis were returning from their trip to Asia. Jay Forrester had promised to teach the Club of Rome all about what he called systems dynamics. Now he needed help from his small department's best computer scientists to demonstrate their work in action. That scientist was, of course, Dennis Meadows. With a few weeks yet before Harvard would open its doors, Dana decided to accompany him, initially planning on listening to what sounded simply like an intriguing seminar. Jay had transcribed the model that he had started on those Swiss napkins in Bern. He now proceeded to present to Aurelio and the executive members of the Club of Rome his model of the world system. It comprised five core elements population, food, industry, resources and pollution and a host of others that interacted with these and each other like education, healthcare, investment and technology. By drawing a series of lines and connectors Jay showed how changes in one part of the system might impact the rest. Jay put forward that it was only with this complete understanding of the system that one could hope to solve the problematique. After two packed weeks of seminars, the Club of Rome was finally satisfied. They offered Jade two million Deutschmarks, around a quarter of a million dollars, donated by the Volkswagen Foundation. It would be enough to kit out an entirely new lab dedicated to systems dynamics, with the latest and most powerful mainframe computers, and to hire over a dozen scientists. But there was a catch. Aurelio, who was now a grandfather, wanted results quickly, within just one year. After that, the Club of Rome intended to set up a permanent think tank in Switzerland to create policy recommendations for the world's governments and help them to implement them. For reasons that Jay never made clear, but were likely influenced by the prospect of the work moving full-time to Switzerland, he declined Aurelio's generous offer to lead the project. Dennis, who had been working side by side with Jay, went home that evening and wrote a proposal. He would use Jay's prototype to show what would happen if current trends continued, and create and test alternative scenarios that could help find solutions to mankind's interconnected problems. As he discussed his ideas with Dana, she decided there and then to give up her hard-earned Harvard Fellowship to help Dennis work on the project. And though her skills would prove valuable in ways she could not yet imagine, to avoid any accusation of nepotism, she insisted on joining the project without pay. Although he was just 28 years old at the time, the Club of Rome accepted Dennis' proposal. Now he and Dana found themselves conducting the first major attempt to model the world system. Jay Forrester had provided the basic structure. Their job was to test it and track down the best numbers to create the scenarios. Over the following months they met with leading geologists, agronomists, chemists, physicists, ecologists, demographers, economists. They studied soil erosion, ozone layer depletion, chemical pollution, acid rain, infant mortality, poverty, malnutrition. They learned about the Earth's mineral deposits and fossil fuel reserves and the energy required to extract them. And they studied the prospects for alternative energy sources like hydrogen, nuclear fusion and solar power. They met with various experts from the likes of the United Nations Population Division, the World Bank, the National Academy of Sciences and the newly formed National Oceanic and Atmospheric Administration, whose scientists presented new and worrying evidence about increasing carbon dioxide emissions from the burning of fossil fuels and how they might lead to dangerous global warming. The team then set about using their data to build their world model. An important feature of the world system is that it is finite. There is a finite amount of land, of mineral deposits and basic elements like oxygen, which we humans and other species depend on. The Kaibab Plateau in Arizona is another such finite system. An elevated area bounded by steep cliff drops on all sides, it is almost impossible for land animals to migrate in or out unaided. Until the late 1800s, the Kaibab was a thriving, balanced ecosystem consisting of deer and natural predators such as wolves and coyotes. That is, until cattle ranchers moved in, leading to a drop in native deer numbers. In an attempt to protect the deer, the government allowed hunters to kill the native predators. Recall how in systems dynamics, Jay Forrester used the term feedback to describe the reaction of one part of a system to changes elsewhere in the system. Negative feedbacks balance or counteract the change, while positive ones reinforce them. Here is Dana using the language of systems dynamics to explain what happens when the predators are removed. The predation rate is part of a negative feedback loop. By the time the predator population comes down to zero, the balance between the positive feedback and the negative feedback is destroyed and starts generating an exponentially growing deer population. As the food gets depleted, it takes longer and longer for it to regenerate. What happens is it drags the deer population down with it. In other words, the removal of the predators, instead of protecting the deer, had the exact opposite effect. It allowed the deer to multiply to such an extent that they eroded their own habitat. The kaibab is an example of what happens when we reach the limits of a finite system. As Dana would later write, when growth is exponential, limits are reached surprisingly quickly. The most common pattern is one of overshoot, beyond the carrying capacity of the environment, followed by collapse. The MIT team's research indicated that in the post-war era, and particularly in the United States, various indicators of human activity also growing exponentially, as Dennis Meadows would later describe when presenting their findings.