Technological Soothsaying
Learning to Live with Uncertainty
In 1943, Thomas Watson, the Chairman of IBM, famously predicted that “there is a world market for maybe five computers.” Even for the room sized super computers of which he spoke, his prediction was a dramatic understatement. And how could he possibly have foreseen the miniaturization and personalization of computer that began to occur some 30 years later? Despite running the world’s foremost computer company, his prediction was wildly inaccurate.
More recently, David Pogue of the New York Times wrote in 2006 that, “everyone’s always asking me when Apple will come out with a cell phone. My answer is, ‘probably never.’” Of course a year later, in 2007, Apple released the first iPhone, almost single-handedly ushering in the age of smartphones. Despite writing about about technology since 1988, Pogue was almost hilariously mistaken here.
Neither Watson nor Pogue are dummies. That isn’t why their predictions were so wrong. The thing is, predicting the future of technological development is often not better than a crapshoot. You might as well make an ad lib and throw darts to fill it in. Even the smartest people with the most expert knowledge in the specific field in question are wrong about their predictions, even near term predictions, at least as often as they are right. But these poor predictions are more than just funny anecdotes, a way to laugh at those much smarter than ourselves and say, “see, even they make mistakes!”
After WWII, the United States had to decide if they were going to pursue a jet air force. Although the aeronautical community was firmly on the path to implementing jet engines in all of their designs, politicians and even some military leaders had yet to be convinced. Jets had substantially higher fuel requirements, and their use by Germany in WWII had been ineffectual, despite their clear technical superiority. As it turns out, by the time politicians were debating the issue, technical experts had pretty much made up their minds. But when the decision was being made, some very powerful people were making some very bad predictions about how useful jet engines were going to be.
It very much matters what people think is going to happen to technology, and those poor predictions can have impacts on what technologies we enjoy and which we miss out on. Sometimes we plunge blindly forward, investing time and resources into technologies that end up going nowhere. Other times we ignore promising technologies, delaying their benefits for years or even decades.
The ideas, the preconceived notions, we have about technology feed back into the decision-making process to, in turn, change what gets developed and what gets ignored.
In the 1920s, a material revolution in flight was getting underway. Although the Wright brothers had built the first airplane primarily out of wood less than 20 years earlier, the technology had advanced in leaps and bounds. According to the National Advisory Committee for Aeronautics (NACA) annual report in 1920, “Wood…is most unsatisfactory…in the future all large airplanes must necessarily be constructed of metal.”
Metal was supposed to have several advantages over wood as an aeronautical material. Metal is not flammable while wood is, thus making metal more fire safe. Metal’s like aluminum and titanium have much higher weight efficiencies than wood. The ability to mass manufacture metal makes it a cheaper option than wood. Finally, metal is a stronger and more durable material than wood.
Yet after over a decade, only 5% of airplanes were made of metal. If it was so superior, why wasn’t it taking off? Because it wasn’t actually as superior as it might initially seem. The danger from fire comes from fuel, not the material the plane is made out of. So metal planes don’t actually improve fire safety. During those 10 years, metal planes were also a quarter to a third heavier than their wooden counterparts. They were also up to twice as expensive to manufacture. And, while expensive materials like aluminum and titanium were indeed more durable, steel was the primary metal used for planes during this time, which was susceptible to corrosion and embrittlement over time.
Metal airplanes weren’t catching on because the performance just didn’t justify it. But engineers in the U.S. and Europe forged ahead nevertheless. Clearly, they weren’t driven by the foresight that one day metal would prove its worth. They were driven by the idea that airplanes should use materials that were themselves the pinnacle of scientific advancement. Scientific advancements in metal materials and their use in industry led engineers to associate metal with progress, and wood as archaic and out of date.
Of course with the benefit of hindsight, we know that metal became the standard material for aviation. But even during WWII, when military airplanes were almost exclusively metal, wooden airplanes could still be dominant. The de Havilland Mosquito was put into service in 1941. It was made of wood, not metal, a trait for which the whole project was nearly canceled. Despite this prejudice against its material construction, the Mosquito was one of the fastest airplanes in operation at the time. It was so successful that it stayed in service until the 1950s.
By the time of the Mosquito, wood technology had advanced a bit beyond just joining raw timbers. The Mosquito was constructed primarily out of laminated construction. Wood fibers are bathed in resin to create thin sheets, which are then stacked in layers and joined with special glue. Material and aviation engineers today will likely recognize this as a description close to describing modern composite materials. Today graphite or carbon fibers, glass fibers, or aramid (kevlar) fibers serve as the foundation for advanced composite materials much the same as wood fibers did for the Mosquito. And these very composite materials have begun to usurp the dominance of metal in aeronautical materials.
Consider that for decades in the early 20th century aeronautics engineers focused nearly all of their time and resources trying to get metal to outperform wood. And yet, despite all the time and attention necessary to make metal work, the engineers still working on the margins of the field still managed to develop with wood the basic material process for composites that have become the primary material for aeronautics today. Did we potentially lose 100 years of progress on composite materials because some engineers thought that metal was more futuristic?
Without being able to compare our world to one that followed an alternative timeline where engineers made a different choice, it isn’t possible to say. Moreover, we can’t rule out that other advancements that came from this obsessive development of metal materials weren’t worth the delay in composite materials. Aluminum and titanium, for example, have been integral in the transformation of aeronautics into aerospace.
The point isn’t that aeronautics engineers made a mistake when deciding that metal was the future of aircraft material. The point is that sometimes even in hindsight it isn’t clear what the best choice was. Technological choice is complex. We almost never have enough information to accurately predict what the outcomes of our choices about technological development will be. Sometimes inferior technologies get locked into markets with increasing returns. Sometimes we bet on a losing technology. Predictions about technologies are barely more than random guesses most of the time. And thus relying on those predictions to determine policy for technological development is probably a losing proposition as well.
But what else are we to do? Or perhaps the better question is, if we must live with uncertainty, what is an alternative to analytical prediction as the basis for decision-making? Perhaps it is better to treat technological development more like an experiment, where the outcomes are revealed through carefully controlled testing rather than predicted beforehand?
Stay tuned for a future post on how such a system might work! In the meantime, I’m curious to hear your thoughts as well!
 | A guest post by
|