IEEE Technology and Society Magazine - March 2022 - 12

Opinion
resources until we are motivated to perform a task.
Distributing cognitive processing to mediating technologies
facilitates this process: offloading seemingly
trivial tasks such as reminders of important
dates, calculations, and information queries.
Dark patterns of design
Reliance on technology can be potentially costly.
The rapid spread of misinformation and disinformation
in our time demonstrate this quite clearly.
A major source of this problem can be attributed
to basic features of memory. When presented with
information, we tend to store it and its source as
separate representations, that is, a failure of source
monitoring [27]. By failing to attend to the source,
disinformation can be stored along with information,
making it difficult to distinguish the good penny from
the bad penny. And bad pennies always turn up. This
can lead to " sleeper effects, " such that information
that is neglected now can influence our later judgments
[32]. Even when a computer is the source,
studies have demonstrated that we can confuse its
productions with our own [29]. Search engine use
can also create the same effect: when search engines
are available, we tend to recall less information ourselves
while still assuming that we possess this knowledge,
that is, the " Google Effect " [30], [31].
These insights can be used for good or ill.
Designs can intentionally exploit our limited attention
resources. " Dark patterns " reflect the ethical
disaffordances of technology [33], [34], nudging
us to perform actions that benefit others, potentially
at our own expense. They can include hidden
costs, deceptive marketing, questionable testimonials,
or using privacy policies and end-user license
agreements as " click wrap " giving organization
unrestricted use of our information [35], [36]. This
is particularly true of social media applications
that are designed to promote personal disclosure.
Behind the benign interface, users' attention and
preferences are being carefully mapped. With little
warning, data subjects can also quickly lose
control over their identities with sudden changes
in policies, such as a recent TikTok decision over
the collection of face and voice data from its users
illustrates [37].
The Cambridge Analytica scandal revealed the
potential scope of such dark patterns. Using psychographic
techniques that can accurately predict
user attributes [38], Cambridge Analytica attempted
12
to target and nudge users to influence their voting
behavior [39]. Yet, despite public uproar due to perceived
violations of privacy and trust, little changed:
one survey suggested that most users did not change
their behavior on Facebook (58%). Of those that did
change their behavior, only 9.6% deactivated their
account, while only 19% reported making fewer
posts to Facebook and 24.8% indicated that they
were more careful [40]. Of course, nudging is a
cheap trick, only working when attention is limited
[41]. Users might discount these attempts as exerting
little influence on their behavior. For instance,
a recent study revealed two networks of over 2,000
bots that produced over 65,000 tweets had little influence
[42]. We must nevertheless consider sleeper
effects and how they can affect us in the long run,
if only subtly.
Rather than focusing only the tail of the coin, we
must also consider its head. By understanding the
possible failures of attention, designers can help
users invest their attention wisely [43]. Outside
of virtual environments, collision and proximity
warning systems can help direct a driver's attention
at critical decision points [44], [45]. Inside
digital environments, behavioral nudging can be
used to encourage citizens to vote [46], [47] and
provide warnings about sharing sensitive content
or content without reading the article [48], [49].
Twitter has suggested that these practices increase
article reading by 40% [50]. Large-scale efforts are
also necessary. Apple's recent App Tracking Transparency
policy was directed toward making users
more aware of when their information is being
shared, a move that proved to be costly for social
media platforms [51].
Changes in policies and design nudges are only
the first step. Users must be empowered through
education and knowledge translation. They must
understand the economics of attention to halt the
inflationary processes that define big data.
 References
[1] M. H. Goldhaber, " The attention economy and the net, "
First Monday, vol. 2, Apr. 1997, doi: 10.5210/fm.v2i4.519.
[2] G. L. Ciampaglia, A. Flammini, and F. Menczer, " The
production of information in the attention economy, "
Sci. Rep., vol. 5, no. 1, pp. 1-6, Aug. 2015.
[3] W. Thorngate, " On paying attention, " in Recent Trends
in Theoretical Psychologsy. New York, NY, USA:
Springer-Verlag, 1988, pp. 247-263.
IEEE Technology and Society Magazine
IEEE Technology and Society Magazine - March 2022

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - March 2022
