Searching for moral reasoning in the IoT

An exploratory analysis of IoT communities and Manifestos

Torey Rubin is one of the thousand customers left in the dark by Emberlight, a start-up that produced a smart socket which used to transform a traditional bulb into a smart light. On the 16th of November 2017, the start-up emailed its customers notifying that it was going out of business because of the pressure from bigger competitors (figure 1).

Figure 1 Emberlight announcement

In the communication sent in a FAQ-like-style, the question that was taking in hostage Torey’s mind (Will my Emberlight Socket continue to work?) got the following answer:

Since Emberlight is an “Internet of Things” product, the device works through 3rd party servers that are hosted in the cloud which have a monthly fee. We do have some credits remaining with our server provider which should enable your devices to work for 3–4 more months, save any major bugs.

Routing information via cloud is a common feature of IoT-related technologies. The cloud affords small businesses and start-ups as Emberlight with straightforward advantages, as for example saving money for those resources-consuming operations as data storage and servers security updates. On the opposite, it comes with some downsides, as experienced by Torey. In light of this, might there have been alternative solutions that could have prevented Torey from being left in the dark? Why Emberlight designers did not opt for the Bluetooth protocol that would have made the socket functioning in local thus not relying solely on a third-party server? Were designers aware of the potential trade-offs they were making while creating the connected device?

Answering these questions required us to map the kind of informal moral reasoning designers undertake when developing IoT products and services.

Why we want to map “moral reasoning” in the IoT domain

When Torey took part to the first round of pre-sales orders Emberlight launched on Kickstarter, he gave little or no account to the design features the socket was made of. He was probably keen about the benefits the smart socket may have guaranteed, either by keeping the bulb separated from the smarts (a beneficial choice in case the bulb breaks or burns out) or by allowing him to make smart those vintage bulbs that would have transformed his living room in a “Pikey Blinders” saloon-like-style. Instead, what probably did not touch Torey’s mind was that the socket design features were the result of the choices and discussions designers undertook while evaluating the best solution to keep together at least two goals: to give customers a functional product and to hit the market.

Plugging objects to the internet, making an ambient sensing and enhancing people with remote control are among the several possibilities afforded by the IoT. Yet its benefits are often marketed with little concern about the possible threats it may bring in terms of unregulated collection and processing of data, lack of minimum security provisions, etc. Understanding these aspects is pivotal as far as new technologies evolve faster than the legislator’s ability to regulate them. For this reason, an approach that looks ex-ante at the consequences (by mapping the moral reasoning of those subjects who develop IoT solutions and services) might help to foresee those downsides that, if left to the mere compliance with the law, will be only tackled ex-post, once they have already spread their effects on society.

We at Virt-EU are deeply concerned with this issue, as we aim to develop a framework that may help to foster ethically aligned design among IoT developers. Thus, to exhaustively map the set of values that potentially drive designers’ actions, our researchers from London School of Economics and IT University of Copenhagen used a set of participatory, ethnographic and interview methods across the events listed in the table below (figure 2). They further enlarged the mapping to IoT Manifestos because “they represent a loud invitation to think in new ways[1]”, delivering an oppositional perspective to the kind of moral reasoning expressed in larger IoT conferences.

Figure 2 — Events

Same issue, a different set of values

The Emberlight issue deals with the mismatch between the physical and the digital part of the product purchased by Torey. The mapping of the informal moral reasoning regarding this problem revealed a different perspective, that stems from the double empirical exercise we undertook.

In major IoT conferences, the issue of “product durability”was framed in terms of “responsibility and design”. The subjects interviewed were mainly concerned about the recourse someone could take if an IoT company goes out of business and the end-user needs regular software updates. Furthermore, we observed that in the case an IoT producer outsources the software development to a third party, the latter claims that if the product brakes, liability will rest on the manufacturer of the hardware and not on who developed the software. According to this perspective, the problem concerning “product durability” is addressed in terms of who is responsible if Torey makes a legal recourse, once the smart socket stops to function. We have interpreted this widespread attitude as the consequence of pressure from market hegemonic forces. For instance, in the case of Emberlight, the choice to opt out for cloud rooting seems to be grounded in the attempt to lower the cost of the smart socket, a necessary condition to survive the tough competition of other start-ups and bigger firms.

Instead, in the analysis of Manifestos, the durability of an IoT product was framed in terms of “sustainability”. Because IoT products are made of hardware and software, usually the lifecycle of physical objects is longer than that of contemporary software, due to the fast rhythm of its continuous development. Several documents call for a re-alignment of digital and physical lifespans, addressing the problem of firmware and software updating that can often make unusable some perfectly functional hardware components. The mismatch between lifespans of hardware and software often leads to intentional obsolescence of products. This kind of strategy looks to Manifestos writers as the consequence of the marketing decision to push an artificial demand for new devices, something they strongly criticize. Instead, every IoT product should guarantee its functionality, as long as the hardware is still working. Following this standpoint, Emberlight designers shouldn’t have relied on cloud rooting as a central feature of their socket and Torey wouldn’t have been left in the dark.

Studying the moral reasoning behind the design of IoT products shows that the same issues may be faced by relying on very different systems of values, something that leads to extremely different solutions.

First insights and future research

The mapping of values in major IoT events returned a picture where the application of the law is viewed as a mean to address ethical issues. Law seems to work as a limit-case for ethics. As a matter of fact, for many IoT developers ethics is articulated with words also appearing in law they must comply with. Only a few of them point to ethical concerns beyond those commonly felt as relevant in the IoT domain, such as privacy and security. This is to say that very few developers are concerned with the problem of product durability faced by Torey because for them a product like the Emberlight socket must be compliant with the laws, more than being designed to serve people needs at best.

Only in a few locales as “ThingsCon Salons”, “The IoT Trustmark” and “The Open IoT Studio Retreat” ethics is enacted as a matter beyond compliance with existing laws, something which is visible in discussions that also occasionally move on legal borders or point towards future laws. These events draw upon a wider perspective on ethics, similar to that coming out from the analysis of manifestos. Indeed, their authors wish to do their part in shaping technological development. In general terms, Manifestos writers seek to pursue a twofold intent: on one side, they suggest future solutions to fix current problems attached to technology, while from another perspective they pose deep concerns about the technology itself. The standpoint emerging from these documents seems to look at the design of IoT products and solutions with an ethical approach, suggesting that people who share it believe that decisions regarding the design of IoT products are the result of a process which must put moral reasoning at the center.

Our mapping has highlighted that a focus on these alternative positions, combined with the data-driven identification of specific sites for sustained involvement within the most active IoT regional hubs across Europe, can help to provide a strategy for formulating our future engagement with this domain. As a matter of fact, we do intend to continue the engagement with these locales, because in the year to come we will enter the places where IoT products are imagined and designed.

In addition, to continue our fieldwork in the two sites we have selected as relevant locifor in-depth domain mapping, namely Amsterdam and London, we will also use the network and legal research to support the identification of specific start-up partners who might become objects for a long-term ethnographic fieldwork.
[1]Parent, M. (2001). The Poetics of the Manifesto. Newness and Nowness. In Manifesto. A Century of Isms. University of Nebraska Press, x–xxxi.