On Technology Addiction
Intro
Technology and social media addiction are significant psychosociological problems that have reduced individual wellbeing for millions and have produced an array of negative economic externalities. Over the past decade, there have been widespread reports of growing addiction to technology and social media in particular with teenagers spending upwards of 7 hours a day on social media on their phones (see link). Various studies have claimed there is a relationship between social media use and increases in depressive symptoms (e.g. Perlis+2021), while others have argued there is no link (e.g. Cunningham+2021). While the psychological effects of technology addiction are debated, its impact on worker productivity is substantial. A recent report in the Harvard Business Review notes that a single person switches between apps and windows more than 3,600 times in a given day. After one is distracted, it takes a substantial amount of time to regain focus (10-20 minutes, i.e. the true cost of multi-tasking) and so many workers are distracted for the majority of their day. Thus, they are less capable of tackling more intellectually challenging tasks, pushing them back until they can really focus, extending timelines and further exacerbating stress surrounding the more difficult tasks, in turn making them appear even more difficult and scarier. Even if one is not using a phone, the mere presence of the device near them looms heavy on the mind, rendering us less capable compared to when we are untethered (Ward+2017). Beyond the impacts on personal productivity, social media usage (driven by addiction) has resulted in increases in political extremism, weakening democracy in the United States (and beyond), and thus presents a clear existential threat to the survival of liberal democracy.
Despite being a serious and recognized concern by many parties (and especially by the creators of technology ; see also "Why Tech Leaders Don't Let Their Kids Use Tech"), technology addiction has become widely normalized and accepted essentially as a necessary consequence of living in an advanced technological society wherein phones and computers are required in order to engage in school, commerce, and socialization. In this essay, I argue that it does not need to be this way and that we can benefit from the incredible capabilities of our technology without them stealing our focus.
Interface design
Most discussions surrounding technology addiction focus on one’s self-control or on the abilities of algorithmically delivered content to stimulate the human brain, but few if any consider the role that the medium itself—the interface of the phone/computer—plays in allowing humans to develop technology addictions. More specifically, the speed and organization of phone/computer interfaces makes it such that it is possible to engage with stimuli with as little effort as possible and with no noticeable delay. User behavior with an interface is itself constrained by the properties of the interface and, because the properties of our technological interfaces make the engagement with addictive stimuli easy, there are more opportunities for users to encounter addictive stimuli and ultimately to develop an addiction. Thus, I argue, that interface design is the key component in enabling addiction to tech-based stimuli, leading it to be significantly more widespread than other behavioral or chemical addictions.
To demonstrate why it is critical to consider the role of the interface, we must first seriously consider what we are asking of users when it is suggested that it is up to the individual to prevent themselves from becoming addicted to technology-based stimuli. Essentially, we are asking users to use one of four solutions: (1) disengage by deleting applications or accounts associated with addicting stimuli, (2) use self-control, (3) regulate display feed algorithms, or (4) alter components of the interface to make it more difficult to access certain applications or to prevent one from using them for too long (e.g. screentime limits).
Regarding solution (1): many of the applications we access addicting stimuli through possess some base level of utility for communicating with friends and family, obtaining news, or learning about the world, and thus it is unreasonable and impractical to request that users delete applications or accounts, ruling out option (1).
Regarding solution (2): self-control is a limited resource (e.g. Muraven and Baumeister 2000, Holfman+2012) and cannot be reliably counted on if the amount of exposure to the temptation is not changed, which is itself unlikely to be reduced given that we need to have our phones readily available for two-factor authentication, communication, mobile banking, GPS navigation, and many other common tasks. Thus it is unlikely that self-control is a serious solution.
Regarding solution (3): Large tech companies which currently benefit from an addicted populace will aim to prevent any serious legislation that could harm their business models. Furthermore, it would likely be extremely difficult to write legislation that contains precise regulatory language such that loopholes would be few if any. Thus, regulations of display feed algorithms seems unfeasible.
Regarding solution (4): Altering the mode of interaction (i.e. the interface) can force users to be more mindful of their activities. This can be done in such a way that users could still easily access the many utilities of their technology but they would experience some level of friction when trying to access applications with addicting content, making them conscious of their habits and empowering them to overcome them in a meaningful way. Some barriers must be imposed to help us use our phones more responsibly. Thus, this solution appears to be an effective middle ground that allows users to benefit from their technology but without being addicted to it.
While solution (4) appears to be a practical path forward, there is a substantial difficulty with its implementation for individuals: alterations of personal phone/computer interfaces require that the individuals themselves have the functional capacity to do so and, furthermore, that the user themselves recognizes their potential vulnerability for developing an addiction. It is easy to see why many users will not possess the functional capacity to alter their interfaces because a certain amount of technical know-how is required and because one must have enough time and energy to do so. While many can satisfy the former requirement, far fewer may feel they have the mental capacity to spare on such a task, especially given that stress levels are much higher now than ever before (see link). Furthermore, the addiction itself may subvert the brain’s motivational system into believing that such a task is too difficult and so would be best avoided in favor of instant stimulation, at least on a subconscious level. Finally, many feel they are not vulnerable or that there is no alternative, i.e. that this amount of available technology-based stimulation is now normal and that no alternative mode of interactivity is possible to imagine, in part because the current landscape of phone and computer interfaces (e.g. desktop environments) are all so similar.
As a result, a plurality of users are functionally incapable of altering their interfaces or do not see the need to and are thus at the mercy of the interface’s defaults, which I have posited specifically enable the development of technology addiction. It is clear that an alternative mode of interactivity is necessary to reduce technology addiction, and I have argued that any individualist solution will fall short. Thus a systematic re-organization of our technological interfaces is necessary and long overdue. To do so properly however, we must first understand what factors go into our development of technology addiction and thereby we must first discuss what addiction itself is so as to prevent it.
What is technology addiction?
To understand technology addiction, it is useful to first understand addiction. Addiction is characterized by cravings and compulsion, a lack of interest in other activities, and by a loss of control (see link). Addiction subverts the reward system in our brains, making it more difficult to enjoy things we used to enjoy. Compulsion to engage with the addictive substance is a matter of anticipation of a reward (e.g. Volkow+2011, Jedras+2013, Hogarth+2020). Addiction is exacerbated by stress, as the engagement with the addictive activity can provide some modicum of relief (e.g. see link). Thus it is sensible to argue that the more stressed someone is, the more susceptible they might be to developing an addiction. It follows then statistically that the portion of a population that suffers from addiction will depend on how stressed the individuals are. We can express this relationship mathematically:
Wherein alpha is some unknown factor ranging between 0 and 1 to account for other potential components (e.g. genetics), sigma is the average stress level of the population which ranges from 0 (minimal) to 1 (maximal; a level greater than 1 would imply an effective dissolution of society), and gamma is an exponent that accounts for likely non-linearity in the relationship between stress and percentage addicted overall.
For the development of addiction to a particular substance, there will be an additional factor which encodes the particularities of the substance’s costs and benefits. This additional factor must be subtracted so as to make the percentage addicted to a given substance less than or equal to the percentage with addiction overall. It must also be a ratio of the costs and the benefits of the particular substance, modulated by some constant to account for the brain’s asymmetric response to costs versus benefits.
Thus, the portion that is addicted to a certain substance (e.g. technology or opioids) can be written as such:
Where STC and STB are the perceived short-term costs (STCs) and benefits (STBs) of using the substance and tau represents the correction factor (>1) that represents the brain’s asymmetric responses to STCs versus STBs. The STC of a substance is essentially what is the amount of effort required to obtain the STB, i.e. a dopamine rush. For a specific individual, the perceived STC and STB of a substance will both depend on the number of prior uses, Nuses , because of the brain’s adaptiveness. Essentially, due to initial lack of familiarity, STC will be higher at lower Nuses and will decrease as Nuses increases as the brain adaptively converts the tasks involved into the subconscious (i.e. moving from system 2 to system 1 in the language of Kahneman).
Now that we have a simplistic model for addiction on the scale of a population, we can apply it to the case of technology addiction. We can assume that the STC and STB are both in the limit of large Nuses and thereby have approached asymptotic values that are fundamental to the case of technology addiction in particular. I posit that these values are set by the parameters of the interface. Given that the STB is primarily well-understood to be some sort of dopamine rush provided by stimulating applications, the STB in the limit of large Nuses can be considered to be some nonzero constant. With this in hand, we must now consider the STC to develop a clear understanding of how technology addiction can be so pervasive and why it is critical to change our technological interfaces so as to overcome it.
The cost of tech-based stimuli
The interfaces through which we access technology-based stimulation have not changed substantially over the last couple of decades. They primarily consist of a landing area where a user can access all their applications (including websites) and files through a small number (<=5) of clicks. This particular design is present in desktop interfaces in the modern operating systems (Windows, MacOS X +iOS, Ubuntu, and Android) that dominate the vast majority of the market share, and this has been the case since their inception. See the below images for a comparison between Windows 95 and Windows 11 and revel in awe at the serious similarity.
The newer one is of course sleeker but they still basically function the same and look the same. When these interfaces were designed, computers were quite slow and the number of files and applications was small. They were typically used to do a single task at a single time. Switching between those tasks required some amount of downtime while waiting for the new application to start. Waiting and boredom are not desirable for the brain, and so the STC of switching applications was perceived as higher. As a result, users spent more of their time on computers in a focused and productive state because they were not constantly switching contexts because their brains knew subconsciously the cost of the boredom that would ensue while waiting passively. As a result, computer applications were viewed primarily as tools to be used and then put away.
However, computing speeds and memory have increased ~1000 times from 1995-2020 and these increases have enabled a marked increase in the number of applications which one can run simultaneously, thus leading to unfocused multitasking because there is no serious penalty for opening more applications without closing others. Additionally, these increases have resulted in there being no perceivable delay when switching between applications or opening new ones. Thus, whenever one faces a substantial cognitive challenge, their brain subconsciously weighs the STCs and STBs of each option: continue with the difficult task (e.g. homework) that requires substantial concerted effort—high STC—or switch to one where stimulation can be achieved with no active effort—minimal STC—such as opening a social media application that algorithmically delivers novel content and which has saved your login information so that you do not have to generate any mindful input (again, limit of high Nuses) . Unsurprisingly, the brain tends to opt for the latter option if no screentime restrictions or barriers are in place, and many end up in a situation where they repeatedly find themselves reopening the same tabs and applications over and over again without consciously intending. As a result, they work only intermittently in breaks between social media based stimulation, substantially dampening their productivity and meaningful engagement with their work.
These problems are further exacerbated by the increase in the number of applications available to users as the average number of desktop applications on a computer has grown at least by a factor of ten (this is a guess as numbers on this are difficult to find) and the number of (active) websites has grown at least by a factor of ~1000 (from 1995 to 2018). Additionally, most of the time that an average professional is working on a computer involves a web browser, and the default configuration of tabs in the major web browsers (Safari, Edge, Chrome, and Firefox) tends to result in users having many (10-100) open, again because there is no penalty for opening more without closing others.
A large (N~50) set of applications becomes unmanageable to organize without some substantial effort on the part of the user and when searching through the available applications to find the one needed to carry out the task at hand, users often find themselves in applications that were not initially intended for but which provide them with easier stimulation. This happens in part because going to your home screen or your list of applications is kind of like walking into a new room (see this for more on this topic for an individual physically relocating). Your brain forgets why you came into the room and reassesses the situation and makes a decision thereafter. You forget why you opened your list of apps and end up on Twitter, Tiktok, Instagram, Gmail, Youtube, Reddit, etc., somewhere where your brain anticipates will stimulate it. This kind of organizational clutter makes us less efficient and thereby makes us more stressed and as a result we are more likely to engage in an addictive activity. This stress is often compounded by stress experienced whilst browsing social media from inflammatory or discouraging content and results in a growing feedback loop of continued use that is typically broken only by extreme disengagement, i.e. deletion of social media accounts or reversion to dumb phones. As outlined earlier, these are not viable options for the masses and are primarily promoted by affluent individuals who have more significant control over their material conditions.
Practically combating technology addiction
In summary, as the speed and memory of computing systems have increased, we have eliminated the barriers between impulses and gratification, thereby minimizing STC to essentially 0 for tech-based stimuli, and as the number of applications/websites has risen, increased clutter and disorganization on computing systems has made us less efficient and more stressed. As a result, the portion of individuals who will be addicted generally will be higher because of higher stress levels, and the majority of those individuals will be addicted to some technology-based stimulation because the psychological cost to access the stimulation is essentially 0, maximizing equation 2.
Simply put, the lack of obstacles for obtaining stimuli on our phones and computers makes it incredibly difficult for the most basic users to use common applications without avoiding technology addiction. And because they are functionally incapable to alter the interfaces that enable the addiction, they are left largely powerless. They end up more strained personally and professionally and are ultimately worse off as a result. With no alternative in sight, it appears hopeless, but humanity has always rearranged the environment to suit its needs (for better or worse), and we will do the same with our digital environment such that it will better suit our own needs rather than those of the large tech companies who profit from capturing our attention.
Much of the work in the past several decades has been in eliminating any potential barriers when using technological interfaces, but the unforeseen consequence of this unprecedented elimination of STC has been tremendously detrimental to the fabric of society and the mental wellbeing of individuals. The good news, however, is that there is a path forward now that we have established that the lack of friction in technological interfaces leads necessarily to technology addiction.
Essentially, we need to introduce some level of friction back into our interfaces so as to increase the STC of the engagement with addictive stimuli. This must be done in such a way that we can still use our phones and computers effectively and quickly without it being so annoying (i.e. nothing like two-factor authentication) that the user will simply revert to the interfaces that are entirely frictionless. How exactly that friction manifests is somewhat uncertain and will likely be determined through some amount of trial and error. Despite the uncertainty, it is clearly imperative that we work towards solving this problem.
TL;DR
The proliferation of technology addiction in modern society is a result of the lack of friction present in modern technological interfaces. These interfaces allow users to access stimulation without any conscious effort and so addiction to some tech-based stimuli becomes effectively inevitable, in part because stress levels are at an all-time high. Users are largely incapable of overcoming the limitations of the interfaces due to a variety of technical or psychological reasons, and so it appears that a systematic adjustment to the structure and organization of technological interfaces is necessary so as to substantially reduce the rate of technology addiction. Re-introducing some level of friction can solve this problem but must be implemented carefully so as to be as user-friendly as possible. Technology addiction is itself an existential problem that poses a great threat against the health of our society and it needs to be overcome in order to move forward as a society so that we can more effectively tackle other existential challenges such as climate change.