“Can’t you just write a program to do this?” is a half-joking question I’m sure most of my computer science peers have received countless times. As computing permeates more and more of our world, we have less of a choice about how it affects us. In order to manage the increasing externalities, a culture shift is required. For too long, we have been operating under the presumption of “good” with regards to new technology. With the magic touch of tech things suddenly become just plainly better. That’s how we get a fridge that can tweet. It’s obviously better than a fridge that can’t tweet!
Unfortunately, this phenomenon isn’t limited to misguided retailers. Software-based elections are a perfect example of the potential for serious problems when this ideology is applied. Elections have been shifting away from paper ballots and towards electronic voting machines. This all changed with the “hanging chad” debacle of 2000. The events of that election happened to coincide with the technology boom of the early 2000s and resulted in a rush to computer voting machines with the promise of increased clarity and efficiency. After all, it was the 21st century — why were we still relying on the cumbersome pen and paper to record our votes? For all the guarantees of computer voting machines, they’ve turned out to be quite the letdown. As Alex Halderman, a professor of Computer Science and Engineering at the University of Michigan, has shown repeatedly, they are vulnerable to a host of attacks and failure states that render them far inferior to paper-based voting systems.
We tend to think of technology only in its perfect state. When everything is running smoothly, computer voting systems sound like a more convenient and efficient way to do things. It is only when we consider the possible imperfections that we can obtain a truly informed assessment of the system. Alas, caution is woefully scarce in many discussions involving technology. Those who are employing the technology fear being seen as old-fashioned, while those building the technology too often take a position similar to Facebook’s former motto, “Move fast and break things.” Risks and externalities are ignored for the creation of a shiny new system that demonstrates “progress.”
But if the product is faulty, won’t the market reflect that and correct itself? The company will make a change or consumers will turn to other options? Not necessarily, as technology failures tend to not make themselves obvious like a faulty air-bag. Your fridge could appear to be functioning perfectly, happily tweeting along, all the while being remotely accessed and enlisted in what is known as a botnet, which is a network of private computers infected with malicious software.
Furthermore, as more nontraditional technology entities utilize software systems, they are frequently dealing with small, external teams of developers that sometimes fail to follow best practices and often lack the resources of a large tech company to dedicate to issues of consumer protection such as security. Security threats tend to be like a game of whack-a-mole — you patch one issue and two more pop up. Without a full-time staff of software professionals, these systems are bound to be one step behind.
Many of these problems stem from the idea that the spread of computing to anything and everything is just and inevitable. It is technology’s version of manifest destiny. I agree that the advent of computing has brought, and will continue to bring, countless powerful improvements to the world. That’s why I am studying computer science. However, we need to think carefully about where and how we use that technology, especially when it’s involved in economies of scale. One tweeting fridge is innocent, a conversation starter even. But one million tweeting fridges potentially becomes a massive botnet, enabling attackers to wreak havoc.
Entities considering the implementation of computing systems should consider worst-case societal outcomes and consult security experts when making their decision. In the world of government, this is a realistic goal and should be expected. In the retail world, this may be unrealistic. Ideally, we would trust these businesses to cast aside technology proposals that serve only as marketing ploys and offer no real consumer benefit. However, this seems to be a social dilemma for which regulation may be needed to induce cooperation. Requiring a product with computing capabilities and large sales to have full-time software staff dedicated to its maintenance is one possibility.
The future of technology is bright. Every day, exciting and meaningful things are being done with the help of computing systems. However, as more and more newcomers enter the computing gold rush, caution will be required in order to avoid fool’s gold.
Chand Rajendra-Nicolucci can be reached at chandrn@umich.edu.