Although Elon Musk's recent cryptic tweets about getting approval to build a Hyperloop system connecting New York and Washington DC are likely to be well received among techno-enthusiasts--many of whom see him as Tony Stark incarnate--there are plenty of reasons to remain skeptical. Musk, of course, has never shied away from proposing and implementing what would otherwise seem to be fairly outlandish technical projects; however, the success of large-scale technological projects depends on more than just getting the engineering right. Given that Musk has provided few signs that he considers the sociopolitical side of his technological undertakings with the same care that he gives the technical aspects (just look at the naivete of his plans for governing a Mars colony), his Hyperloop project is most likely going to be a boondoggle--unless he is very, very lucky.

Don't misunderstand my intentions, dear reader. I wish Mr. Musk all the best. If climate scientists are correct, technological societies ought to be doing everything they can to get citizens out of their cars, out of airplanes, and into trains. Generally I am in favor of any project that gets us one step closer to that goal. However, expensive failures would hurt the legitimacy of alternative transportation projects, in addition to sucking up capital that could be used on projects that are more likely to succeed. So what leads me to believe that the Hyperloop, as currently envisioned, is probably destined for trouble?

Musk's proposals, as well as the arguments of many of his cheerleaders, are marked by an extreme degree of faith in the power of engineering calculation. This faith flies in the face of much of the history of technological change, which has primarily been a incremental, trial-and-error affair often resulting in more failures than success stories. The complexity of reality and of contemporary technologies dwarfs people's ability to model and predict. Hyman Rickover, the officer in charge of developing the Navy's first nuclear submarine, described at the length the significant differences between "paper reactors" and "real reactors," namely that the latter are usually behind schedule, hugely expensive, and surprisingly complicated by what would normally be trivial issues. In fact, part of the reason the early nuclear energy industry was such a failure, in terms of safety oversights and being hugely over budget, was that decisions were dominated by enthusiasts and that they scaled the technology up too rapidly, building plants six times larger than those that currently existed before having gained sufficient expertise with the technology.  

Musk has yet to build a full-scale Hyperloop, leaving unanswered questions as to whether or not he can satisfactorily deal with the complications inherent in shooting people down a pressurized tube at 800 miles an hour. All publicly available information suggests he has only constructed a one-mile mock-up on his company's property. Although this is one step beyond a "paper" Hyperloop, a NY to DC line would be approximately 250 times longer. Given that unexpected phenomena emerge with increasing scale, Musk would be prudent to start smaller. Doing so would be to learn from the US's and Germany's failed efforts to develop wind power in 1980s. They tried to build the most technically advanced turbines possible, drawing on recent aeronautical innovations. Yet their efforts resulted in gargantuan turbines that failed often within tens of operating hours. The Danes, in contrast, started with conventional designs, incrementally scaling up designs andlearning from  experience. 

Apart from the scaling-up problem, Musk's project relies on simultaneously making unprecedented advances in tunneling technology. The "Boring Company" website touts their vision for managing to accomplish a ten-fold decrease in cost through potential technical improvements: increasing boring machine power, shrinking tunnel diameters, and (more dubiously) automating the tunneling process. As a student of technological failure, I would question the wisdom of throwing complex and largely experimental boring technology into a project that is already a large, complicated endeavor that Musk and his employees have too little experience with. A prudent approach would entail spending considerable time testing these new machines on smaller projects with far less financial risk before jumping headfirst into a Hyperloop project. Indeed, the failure of the US space shuttle can be partly attributed to the desire to innovate in too many areas at the same time. 

Moreover, Musk's proposals seem woefully uninformed about the complications that arise in tunnel construction, many of which can sink a project. No matter how sophisticated or well engineered the technology involved, the success of large-scale sociotechnical projects are incredibly sensitive to unanticipated errors. This is because such projects are highly capital intensive and inflexibly designed. As a result, mistakes increase costs and, in turn, production pressures--which then contributes to future errors. The project to build a 2 mile tunnel to replace the Alaska Way Viaduct, for instance, incurred a two year, quarter billion dollar delay after the boring machine was damaged after striking a pipe casing that went unnoticed in the survey process. Unless taxpayers are forced to pony up for those costs, you can be sure that tunnel tolls will be higher than predicted. It is difficult to imagine how many hiccups could stymie construction on a 250 mile Hyperloop. Such errors will invariably raise the capital costs of the project, costs that would need to be recouped through operating revenues. Given the competition from other trains, driving, and flying, too high of fares could turn the Hyperloop into a luxury transport system for the elite. Concorde anyone? 

Again, while I applaud Musk's ambition, I worry that he is not proceeding intelligently enough. Intelligently developing something like a Hyperloop system would entail focusing more on his own and his organization's ignorance, avoiding the tendency to become overly enamored with one's own technical acumen. Doing so would also entail not committing oneself too early to a certain technical outcome but designing so as to maximize opportunities for learning as well as ensuring that mistakes are relatively inexpensive to correct. Such an approach, unfortunately, is rarely compatible with grand visions of immediate technical progress, at least in the short-term. Unfortunately, many of us, especially Silicon Valley venture capitalists, are too in love with those grand visions to make the right demands of technologists like Musk. 

When reading some observer's diagnoses of what ails the United States, one can get the impression that Americans are living in an unprecedented age of public scientific ignorance. There is reason, however, to wonder if people today are really any more ignorant of facts like water boiling at lower temperatures at higher altitudes or if any more people believe in astrology than in the past. According to some studies, Americans have never been more scientifically literate. Nevertheless, there is no shortage of hand-wringing about the remaining degree of public scientific illiteracy and what it might mean for the future of the United States and American democracy. Indeed, scientific illiteracy is targeted as the cause of the anti-vaccination movement as well as opposition to genetically modified organisms (GMOs) and nuclear power. However, I think such arguments misunderstand the issue. If America has a problem with regard to science, it is not due to a dearth of scientific literacy but a decline in science's public legitimacy.
 
The thinking underlying worries about widespread scientific illiteracy is rooted in what is called the “deficit model.” In the deficit model, the cause of the discrepancy between the beliefs of scientists and those are the public is, in the words of Dietram Scheufele and Matthew Nisbet, a “failure in transmission.” That is, it is believed that negligence of the media to dutifully report the “objective” facts or the inability of an irrational public to correctly receive those facts prevents the public from having the “right” beliefs regarding issues like science funding or the desirability of technologies like genetically modified organisms. Indeed, a blogger for Scientific American blames the opposition of liberals to nuclear power on “ignorance” and “bad psychological connections.” It is perhaps only a slight exaggeration to say that the deficit model depicts anyone who is not a technological enthusiast as uninformed, if not idiotic.
 
Regardless of whether or not the facts regarding these issues are actually “objective” or totally certain (both sides dispute the validity of each other’s arguments on scientific grounds), it remains odd that deficit model commentators view the discrepancy between scientists’ and the public’s views on GMOs and other issues as a problem for democracy. Certainly they are correct that it is preferable to have a populace that can think critically and suffers from few cognitive impairments to inquiry when it comes to wise public decision making. Yet, the idea that, when properly “informed” of the relevant facts, scientifically literate citizens would immediately agree with experts is profoundly undemocratic: It belittles and erases all the relevant disagreements about values and rights. Such a view ignores, for instance, the fact that the dispute over GMO labeling has as much to do with ideas about citizens’ right to know and desire for transparency as the putative safety of GMOs. By acting as if such concerns do not matter – that only the outcome of recent safety studies do – the people sharing those concerns are deprived of a voice. The deficit model inexorably excludes those not working within a scientistic framework from democratic decision making.
 
Given the deficit model’s democratic deficits as well as the lack of any evidence that scientific illiteracy is actually increasing, advocates of GMOs and other potentially risky instances of technoscience ought to look elsewhere for the sources of public scientific controversy. If anything has changed in the last decades it is that science and technology have less legitimacy. Indeed, science writers could better grasp this point by reading one of their own. Former Discover writer Robert Pool notes that the point of legal and regulatory challenges to new technoscience is not simply to render it safer but also more acceptable to citizens. Whether or not citizens accept a new technology depends upon the level of trust they have of technical experts (and the firms they work for). Opposition to GMOs, for instance, is partly rooted in the belief that private firms such as Monsanto cannot be trusted to adequately test their products and that the FDA and EPA are too toothless (or captured by industry interests) to hold such companies to a high enough standard. Technoscientists and cheerleading science writers are probably oblivious to the workings and requirements for earning public trust because they are usually biased to seeing new technologies as already (if not inherently) legitimate.
 
Those deriding the public for failing to recognize the supposedly objective desirability of potentially risky technology, moreover, have fatally misunderstood the relationship between expertise, knowledge, and legitimacy. It is unreasonable to expect members of the public to somehow find the time (or perhaps even the interest) to learn about the nuances of genetic transmission or nuclear safety systems. Such expectations place a unique and unfair burden on lay citizens. Many technical experts, for instance, might be found to be equally ignorant of elementary distinctions in the social sciences or philosophy. Yet, few seem to consider such illiteracies to be equally worrisome barriers to a well-functioning democracy. In any case, as political scientists Joseph Morone and Edward Woodhouse argue, the position of the public is not to evaluate complex or arcane technoscientific problems directly but to decide which experts to trust to do so. Citizens, according to Morone and Woodhouse, were quite reasonable to turn against nuclear power when overoptimistic safety estimates were proven wrong by a series of public blunders, including accidents at Three Mile Island and Chernobyl, as well as increasing levels of disagreement among experts. Citizens’ lack of understanding of nuclear physics was beside the point: The technology was oversold and overhyped.  The public now had good grounds to believe that experts were not approaching nuclear energy or their risk assessments responsibly. Contrary to the assumptions of deficit modelers, legitimacy is not earned simply through technical expertise but via sociopolitical demonstrations of trustworthiness.
 
If technoscientific experts were to really care about democracy, they would think more deeply about how they could better earn legitimacy in the eyes of the public. At the very least, research in science and technology studies provides some guidance on how they ought not to proceed. For example, after post-Chernobyl accident radiation rained down on parts of Cumbria, England, scientists quickly moved in to study the effects as well as ensure that irradiated livestock did not get moved out of the area. Their behavior quickly earned them the ire of local farmers. Scientists not only ignored the relevant expertise that farmers had regarding the problem but also made bold pronouncements of fact that were later found to be false, including the claim that the nearby Sellafield nuclear processing plant had nothing to do with local radiation levels. The certainty with which scientists made their uncertain claims as well as their unwillingness to respond to criticism by non-scientists led farmers to distrust them. The scientists lost legitimacy as local citizens came to believe that they were sent there by the national government to stifle inquiry into what was going on rather than learn the facts of the matter.
 
Far too many technoscientists (or at least their associated cheerleaders in popular media) seem content to repeat the mistakes of these Cumbrian radiation scientists. “Take your concerns elsewhere. The experts are talking,” they seem to say when non-experts raise concerns, “Come back when you’ve got a science degree.” Ironically (and tragically), experts’ embrace of deficit model understandings of public scientific controversies undermines the very mechanisms by which legitimacy is established. If the problem is really a deficit of public trust, diminishing the transparency of decisions and eliminating possibilities for citizen participation is self-defeating. Anything looking like a constructive and democratic resolution to controversies like GMOs, fracking, or nuclear energy is only likely to happen if experts engage with and seek to understand popular opposition. Only then can they begin to incrementally reestablish trust. Insofar as far too many scientists and other experts believe they deserve public legitimacy simply by their credentials – and some even denigrate lay citizens as ignorant rubes – public scientific controversies are likely to continue to be polarized and pathological.