ArticlePDF Available

The Digital Phenotype: a Philosophical and Ethical Exploration

Authors:

Abstract and Figures

The concept of the digital phenotype has been used to refer to digital data prognostic or diagnostic of disease conditions. Medical conditions may be inferred from the time pattern in an insomniac’s tweets, the Facebook posts of a depressed individual, or the web searches of a hypochondriac. This paper conceptualizes digital data as an extended phenotype of humans, that is as digital information produced by humans and affecting human behavior and culture. It argues that there are ethical obligations to persons affected by generalizable knowledge of a digital phenotype, not only those who are personally identifiable or involved in data generation. This claim is illustrated by considering the health-related digital phenotypes of precision medicine and digital epidemiology. Full text avaiable: https://rdcu.be/V1qH
This content is subject to copyright. Terms and conditions apply.
This is only a summary version of the paper, including the introduction and
conclusion (full text) and a bulletpoint summary of the other sections. Courtesy of
Springer, the full paper is provisionally available here https://rdcu.be/V1qH
The digital phenotype: a philosophical and ethical exploration

              
          
              


          
     
    
    !          
      

    "        
     
  #        

Introduction
     
                
       
$%&'()*+
     
      $%  &'()*  
             $,  &'(-*  !  
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33

#(
                        
                  
              .    
                  
        
/
               
                

 

              
   #     
 
&0
  #      $      
*"

  
1
!(!"

1 !234
.&'(-5$#4*62#
.#
78


/+
2bots$*

$&'(9*#:,+;1:
0"
;7.,:77
<!;Information, Communication & Society())
$%(&'(&*1=)'
2
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
1
2
3
4
5
6
7
8
9
10
11
12
13
14
.+ !  
          #
                  
       !
    
    
   ! & !
                    
1
                        #
!
#>!?!
                  #
     

 +       1 
                 digital
phenotype                   
#
1. What is the digital phenotype?
3
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
Figure 1. Analogies between the extended phenotypes of spiders and humans. Built with Apple
Keynote 8.0.1, except for web image, by Denis Frezzato, from The Noun Project, CC BY 3.0,
https://commons.wikimedia.org/w/index.php?curid=24126123
Organisms evolve together with the environments they shape or create. E.g.
beehives, beaver dams.
Digital data is analogous to beaver dams or beehives in humans: these are all
collectively created extended phenotypes involved in evolutionary feedback loops
(Richard Dawkins).
Human evolution is also cultural. Data affects the beliefs, norms and behavior in the
human population, thus affecting the cultural dimension of human evolution, most
directly (effects on other levels of evolution cannot be excluded in principle).
The metaphor of data as “digital footprint” is misleading. Rather data for humans is
like the beehive for bees, not a passive trace left somewhere, but something that
groups create collectively and in a way that influences the further behavior of its
producers
2. The ethics of the digital phenotype
2.1 The limits of the personal data protection approach
The data protection approach protects the interests of those who are identifiable in
the digital phenotype.
4
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
This fails to consider the interests of those who are not identifiable, that is, all other
people who may also be affected when a new digital phenotype is created.
Every time generalizable knowledge is produced by studying a digital phenotype,
there are people who are affected outside those identifiable in the data. E.g. we
discover that smoking causes cancer. This affects all smokers (e.g. higher insurance
prices for all smokers).
N.B. The fact that generalized knowledge harm some people all things considered
may be a reason against producing it; but it may also not be, and in many cases, it is
not, an overriding reason and a reason all things considered for not producing such
knowledge. The benefit for all may outweigh the harm of the few.
N.B.2. The problem of generalized knowledge has been recognized in the past for
genetic data. Many have claimed that genetic data is special because it is shared
among relatives. That is a red herring. If knowledge produced with data from person
A allows inferences about person B, then knowledge production about A creates
potential risks (e.g. privacy and discrimination risks) for person B. That is obviously
the case when you produce genetic knowledge about A and B is A’s twin. But it is
also true in the smoking causes cancer case, even if there are less inferences from A
to B to be made.
2.2 The limits of the libertarian approach
The libertarian approach cares about the interests of the people involved in the
production of the digital phenotype, irrespective of whether or not they are
identifiable. This view has several problems:
oPeople can be involved in data production in two different ways: as persons
causing the data collection process to exist and as persons causing the
particular data to exist. E.g. as a shop owner I tape all my clients (person
causing the data collection process to exist); as a client I am taped by the
shop owner (I am the cause of some of the data collected). The interests of
parties involved in data production may conflict. The libertarian approach
does not provide principles to solve those conflicts (except that people are
free to make individual contracts without use of coercion – violence).
oSome digital phenotypes (e.g. Google searches from the beginning of its
existence) are due to the contribution of millions of persons. Shared
ownership to such phenotypes among the co-creators (e.g. all persons typing
a search string on Google) dilutes the practical value of ownership in terms of
control, for each.
3 Governing the health-relevant digital phenotype
In relation to health we can distinguish three levels of analysis: medical data, data
concerning health (in the GDPR sense) and health-relevant digital phenotypes. The
5
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
health-relevant digital phenotype is the broadest of the three concepts,
encompassing the other two. Tweets about vaccination constitute a health-relevant
digital phenotype (it can be studied for public health purposes) but are not medical
data or data concerning health in the GDPR sense
Case study 1. P4 medicine
Personalized medicine using big data generates a digital phenotype that illustrates
the ethical issues for people who are neither identifiable nor involved in data
production.
The big issue for governance is to protect groups that may suffer harm from
discrimination produced by algorithms that are learned from the data.
It may be exceedingly difficult to identify some of these groups.
It may be difficult to use existing governance models for genomic research which
focus on the protection of a specific population, e.g. endangered ethnic minority,
vulnerable to stereotyping, etc…
Case study 2. Digital epidemiology and the use of internet data in public
health
Another example is the digital phenotype from large platforms such as Google. Here the
issue is that these platforms produce data that may potentially benefit populations, e.g. if
they are used for public health.
But the data collection process and algorithms are not optimized for epidemiology.
Such platforms should at least become more transparent about their data and
algorithms, allowing broadened access to their data and algorithms, in order to enable
researchers (not working for or in direct partnership with Google) to criticize and improve
them.
Conclusions


           
!
"
0
                
        "      
@0
6
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
"

"
/?#
 /29
       
      3  
                    #
<

"A
                
 /      
"

Reference list:
/5B.737.
B@8C@2
<7&'(=:7
<D4.+3E4
.;45C&&'(=
1FFF###
######F
/5B.737.
BCG@2&'(=
:33!<.4.6
H/..3
5<4.@;
Croatian Medical Journal)-$&*1&'-I(?
@B%5(JJ9Principles of
Biomedical Ethics96E2E7/
@.,5&'(&:5K@.;
Information, Communication & Society()$)*1==&I-J
@/&''-:!@313
57;Journal of Political Philosophy()$(*1&&I9)
5/@(JJ(:257<1
67;
Business Horizons?9$9*1?JI9L
7
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
5#7BH(JL(Cultural Transmission
and Evolution: A Quantitative Approach2121
2E2
5M/(JJ)Self-Ownership, Freedom, and Equality
515E2
.%%45C<NO,/
@/.24&'(-:/
M1.</2
5!;Big Data & Society9$&*1
&')?J)(-(--&=))9
.<(JJJThe Extended Phenotype: The Long Reach of
the Gene&616E2
35.M4&'(9:E@.2
42<;@135
.M4544
/
3M&''J:!!1
3724!
/752@
!;Journal of Medical Internet Research(($(*1((
B&'((:!C2!;
Minds and Machines&($9*1)9JI==
PPP&'(9:6..2M2;
Philosophy & Technology&-$(*1(
PPP&'(=:64.<
2;Philosophy & Technology&J$9*1?'-I?(&
PPP&'(=:M21/.!;!
Group Privacy: New Challenges of Data Technologies
BB@7L?I(''
517
.&'(9:6!..
575<;Journal of
Broadcasting & Electronic Media)L$(*1)JI-)
M %&''LOn Human Rights616E2
4.%&''LWhen Species MeetQ&&9
1E2
4M &''):57<1/3
;The Geneva Papers on Risk and Insurance -
Issues and Practice?'$?*1?L-I9'J
4B%5B8C.2&'():!
@./456H;
BMC Medicine(?19
%3%B&'')Evolution in Four Dimensions:
Genetic, Epigenetic, Behavioral, and Symbolic Variation in the
History of Life51!2
%74@H2%@4%7
@&'():.2;Nature Biotechnology
??$)*19=&I=?
8
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
%>2%.%BR2B
B&''=:7!331
!/3 7B;
Applied Soil Ecology?&$&*1()?I(=9
,B&'(-:E/!77E;BBC News
(&'(-
1FFFF#?J(&='&-
,2&''(Science, Truth, and Democracy6ACS1
6E2
B.<,M,/Q
&'(9:2M1@./;
Science?9?$=(-=*1(&'?I)
B.&'():42.31/5
5;Health Promotion International?'$(*1(-9IL?
PPP&'(=:.573.1!
..I/;Big Data &
Society?$(*1(I)
PPP&'(=:1B.B.BB
7;Leisure Studies(I?
%T2 &'(=:!23 
7.;!Tenth International AAAI Conference on
Web and Social Media
1FFFFF!5H7F!5H7(=FFF(
?(=?
@&'(-:.4#<!
132M;InformationL$?*
1FFF&'-L#&9LJFLF?F--
PPP&'(-:34#<!1/
C<;Ethics and Information Technology(I(J
@2/7H
B&'(=:3/1
.;Big Data & Society?$&*1&')?J)(-(==-J=-J
C/7&'((Toward Precision Medicine:
Building a Knowledge Network for Biomedical Research and a New
Taxonomy of DiseaseH.51C/2
C<(J-9Anarchy, State, and UtopiaCS1@
@
6&'():6;!The Onlife Manifesto
B-I(?7!2
2B&'(?:DHM+U21
<M;Bioethics&-$-*1?)-I9('
2@/@&'(?:/7#@/
M<@;Medical Law Review&(
$(*1-(IJ(
Regulation on the Protection of Natural Persons with Regard to the
Processing of Personal Data and on the Free Movement of Such
Data, and Repealing Directive 95/46/EC (General Data Protection
Regulation 2016/ 679)&'(=1FF#F#
F3CFVFW
9
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
X16%BY&'(=((J'('''('(3CMZX6%1B1&'(=1((J1
65
<.%2 &'(9:7B7
@;Science?9=$=&(?*1('=?I('=9
7[B@%@..@%
7@5@35
&'(&:.3;PLOS Comput BiolL$-*1(''&=(=
7[7,&'((:/Q
7671!!
..5;PLoS Comput Biol-$('*1
(''&(JJ
Q3 %.B&'(=:4
.3@.;H46&'(=
1FFFFFFF
Q3 7[B5 %7
@&'():35@.2
4;PLoS Comput Biol(($&*1(''?J'9
H4&'(?The Connected Self: The Ethics and
Governance of the Genetic Individual5E2
H .3B@C&''):75
M122D@+;
Genomics, Society and Policy($?*1LI&(
H .3<CH &''?:5
M@;New England Journal of
Medicine?9J$(&*1((L'I((L9
H/#H/&'(=:H/.
35<4.
@;HH/6&&&'(=
1FFF#F#####
######F
\7MM2<V\4
\@S\&'(9:B5.1
7C.;!Eighth
International AAAI Conference on Weblogs and Social Media
1FFFFF!5H7F!5H7(9FFFL
'9&
10
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358

Supplementary resource (1)

... how they behave) characteristics and behaviors (Zarate et al., 2022). As such, a person's actions can be seen as reflections/footprints of their genetic predispositions combined with their past experiences (Dawkins, 1982;Loi, 2019). This implies that how an individual behaves may provide insights and traces into their overall healthincluding their mental health and well-being (Rozgonjuk et al., 2023;Zarate et al., 2022). ...
... anxious and depressive behaviors; Kozybska et al., 2022;Zarate et al., 2022). The methodologies employed to transform this digital data into health insights are communicated as digital phenotype (DP) processes (Loi, 2019). Generally, these processes involve the use of digital devices to gather passive/ objective data without emphasizing the specific technology or data type (Loi, 2019). ...
... The methodologies employed to transform this digital data into health insights are communicated as digital phenotype (DP) processes (Loi, 2019). Generally, these processes involve the use of digital devices to gather passive/ objective data without emphasizing the specific technology or data type (Loi, 2019). However, differentiation of data across collection devices has been recommended to address ecological validity concerns (Zarate et al., 2022). ...
Article
Full-text available
Research has supported that a gamer's attachment to their avatar can offer significant insights about their mental health, including anxiety. To assess this hypothesis, longitudinal data from 565 adult and adolescent participants (Mage = 29.3 years, SD = 10.6) was analyzed at two points, six months apart. Respondents were assessed using the User-Avatar Bond (UAB) scale and the Depression Anxiety Stress Scale (DASS) to measure their connection with their avatar and their risk for anxiety. The records were processed using both untuned and tuned artificial intelligence [AI] classifiers to evaluate present and future anxiety. The findings indicated that AI models are capable of accurately and autonomously discerning cases of anxiety risk based on the gamers' self-reported UAB, age, and duration of gaming, both at present and after six months. Notably, random forest algorithms surpassed other AI models in effectiveness, with avatar compensation emerging as the most significant factor in model training for prospective anxiety. The implications for assessment, prevention, and clinical practice are discussed.
... The concept of the extended phenotype encompasses the observable characteristics resulting from an organism's genes and environment and thus includes the behavioral expressions that arise from an individual's predispositions and life experiences (Dawkins, 1982;Loi, 2019). Therefore, it is assumed that an individual's behavioral phenotype/profile (i.e., how a person behaves) carries valuable information about their health conditions, including potential indicators of mental health (Zarate et al., 2022;Rozgonjuk et al., 2023). ...
... By analyzing digital footprints, researchers aim to derive diagnostic information relevant to an individual's overall health (Insel, 2018). The collection of such methodologies, aiming to translate digital and/or cyber data into health and mental health information about the individual is known as 'digital phenotype' procedures (DP; Loi, 2019). As such, the DP is predicated on the utilization of digital devices and the gathering of objective/passive data, without making distinctions based on the specific technology employed or the nature of the data collected (Zarate et al., 2022). ...
... Expanding the cyber-phenotyping framework to encompass the multidimensional data embedded in the UAB could substantially broaden the lens for deriving mental health insights from gaming activities. Indeed, the UAB meets the criteria for a DP as it manifests in online ecosystems (i.e., the virtual environment), and contains observable patterns reflecting an individual's predispositions and mental state (Loi, 2019). Similar to physical phenotypes offering insights into health, the way users customize, relate to, and interact with their virtual identities offers quantifiable data providing information on psychological well-being (Szolin et al., 2022). ...
... It has also been proposed that the UAB could operate as a form of 'digital phenotype', meaning a digital/gamified footprint of an individual's mental health, that, if analysed, can be translated into information not only concerning the gamer's risk of GD, but also for other psychopathological conditions (e.g., depression, anxiety [Loi, 2019;Stavropoulos et al., 2021;Zarate, Stavropoulos, Ball, de Sena Collier, & Jacobson, 2022]). Despite the consistent associations between GD and the UAB in the extant literature, the translation of the UAB into GD risk has never to date, to the best of the authors' knowledge, been investigated Liew et al., 2018;Ratan et al., 2020). ...
... These interpretations reinforce (and align with) the proposed notion of 'digital phenotype', suggesting that an individual's cyber-behaviour and choices, such as their useravatar customization and bond, may operate as a unique 'footprint' of what they are experiencing offline, if/when appropriately translated (Loi, 2019;Stavropoulos et al., 2021;Zarate et al., 2022). This possibility is additionally strengthened by the work of Lemenager et al. (2020), who reported: (i) a consistent association between disordered gaming and bonding with the avatar, and; (ii) enhanced activation of brain regions during times an individual is consumed by thoughts regarding their avatar. ...
... In particular, from a conceptual perspective, and in relation to the notion of digital phenotype, the use of ML/AI to show the GD diagnostic potential of the UAB, expands past studies in the field, suggesting the need for exploration of further health and mental information likely embedded within the UAB, independent of GD risk (e.g. depression, anxiety; Lemenager et al., 2020;Loi, 2019;. Overall, the present study suggests that GD risk can be predicted using ML/AI algorithms, that are capable of combining different variables on a large scale with reduced rates of misdiagnosis, providing more accurate diagnostic and/or risk indicators. ...
Article
Full-text available
Background and aims Gaming disorder [GD] risk has been associated with the way gamers bond with their visual representation (i.e., avatar) in the game-world. More specifically, a gamer's relationship with their avatar has been shown to provide reliable mental health information about the user in their offline life, such as their current and prospective GD risk, if appropriately decoded. Methods To contribute to the paucity of knowledge in this area, 565 gamers ( M age = 29.3 years; SD =10.6) were assessed twice, six months apart, using the User-Avatar-Bond Scale (UABS) and the Gaming Disorder Test. A series of tuned and untuned artificial intelligence [AI] classifiers analysed concurrently and prospectively their responses. Results Findings showed that AI models learned to accurately and automatically identify GD risk cases, based on gamers' reported UABS score, age, and length of gaming involvement, both concurrently and longitudinally (i.e., six months later). Random forests outperformed all other AIs, while avatar immersion was shown to be the strongest training predictor. Conclusion Study outcomes demonstrated that the user-avatar bond can be translated into accurate, concurrent and future GD risk predictions using trained AI classifiers. Assessment, prevention, and practice implications are discussed in the light of these findings.
... Datenschutz: Die Nutzung von KI kann zu erheblichen Herausforderungen im Datenschutz, wie Datenlecks oder fehlerhaft trainierten Algorithmen, führen (Hagendorff 2020; Chomanski 2021;Morley et al. 2021). Die Datenschutzgrundverordnung (DSGVO) wird dabei als entsprechender rechtlicher Rahmen erachtet(Loi 2019; Saetra und Danaher 2022), wobei der bestehende Schutz personenbezogener Daten möglicherweise nicht ausreicht, um Diskriminierung von nicht identifizierten Personen zu verhindern. Resultierend daraus ergibt sich eine vertiefte Diskussion über die die Wahrung der Privatsphäre, der wissenschaftlichen Gültigkeit und den Eigentumsrechten(Loi 2019). ...
... Die Datenschutzgrundverordnung (DSGVO) wird dabei als entsprechender rechtlicher Rahmen erachtet(Loi 2019; Saetra und Danaher 2022), wobei der bestehende Schutz personenbezogener Daten möglicherweise nicht ausreicht, um Diskriminierung von nicht identifizierten Personen zu verhindern. Resultierend daraus ergibt sich eine vertiefte Diskussion über die die Wahrung der Privatsphäre, der wissenschaftlichen Gültigkeit und den Eigentumsrechten(Loi 2019). Zustimmung: Die ethische Verpflichtung zur Einholung der Zustimmung zur Veröffentlichung anonymisierter Daten besteht, trotz fehlender DSGVO-Vorschriften für anonymisierte oder pseudonymisierte Daten. ...
Article
Full-text available
Zusammenfassung Process Mining (PM) stellt eine wachsende Disziplin dar, die aufgrund ihres Potenzials zur Verbesserung von Geschäftsprozessen immer mehr Aufmerksamkeit von Forschern und Anwendern auf sich zieht. Wie jede neue Technologie gibt es jedoch auch im Kontext von PM-Bedenken hinsichtlich der ethischen Anwendung. Gerade bezogen auf Erhebung, Verarbeitung und Nutzung von Daten kann es hierbei zu Problemen kommen. Dieser Artikel zielt daher darauf ab, anhand einer Literaturanalyse ethische Implikationen im Process Mining herauszuarbeiten. Dabei wurden 39 Artikel aus sechs Zeitschriften im Bereich PM und 24 Artikel aus vier Zeitschriften im Bereich Datenethik analysiert. Die Ergebnisse zeigen das wachsende Interesse an der Datenethik und PM, aber es befasst sich nur ein geringer Anteil der analysierten PM-Artikel mit datenethischen Grundsätzen. Weitere Forschung ist in Bereichen bestimmter datenethischer Grundsätze, wie Datenqualität und der informierten Zustimmung, erforderlich. Insgesamt bietet diese Studie einen Ausgangspunkt für weitere Forschungen zur ethischen Nutzung von Daten bei der Anwendung von PM und verdeutlicht, dass diesem Bereich mehr Aufmerksamkeit gewidmet werden sollte.
... But what exactly is the basis for these findings, how could/would the data from digital footprints be clinically employed, and could DP be justified in practice? Although there has been significant scientific research into DP and critical analysis (including techno-social criticism) from the social sciences (Birk & Samuel, 2020 and ethics/philosophy (Loi, 2019;Mulvenna et al., 2021), there has been only a relatively small set of papers to emerge that offer a focused ethical analysis of DP with respect to its implementation in clinical mental healthcare. ...
Article
Full-text available
Digital phenotyping (DP) refers to the emerging field within digital (mental) health that involves the collection of data from individual’s digital devices (smartphones, wearable, Internet usage, etc.) to monitor and analyse their behaviours, activities and health-related patterns. That such ‘digital footprint’ data can be mined for behavioural insights is an intriguing idea, which has motivated an increasing amount of research activity, particularly in the field of digital mental health. Whilst of potentially revolutionary utility in (mental) healthcare, the idea of DP also raises a set of rich sociotechnical, ethical and philosophical considerations, and a cottage industry of sociotechnical and ethical critiques of DP has emerged, particularly within the humanities. Within this scene of early empirical investigation in the health/behavioural sciences on the one hand and general conceptual critiques from the humanities on the other, in this paper we instead explore a case for the potential utility of DP in circumstances of clinical mental health practice and examine its ethical dimensions in this context. After providing an explicatory framework for DP and making a case for it within this framework, we subsequently examine the ethical pros and cons of three topics pertaining to DP in clinical practice, namely (1) issues in psychometric assessment/testing, (2) the role and responsibilities of mental health practitioners with regard to DP technology, and (3) the value DP can afford clients in terms of self-awareness/empowerment and strengthening the therapeutic alliance with their clinician.
... These studies have identified several opportunities to give workers insight into their wellbeing by leveraging data from technology a worker interacts with-email use [76], smartphones [16,42,81], wearables [40,77,103,115], webcams [58], networked devices [38,91], and social media [9,109,112,113,116]. In a broader sense, the approach of digitizing an individual's free-living behaviors to identify their health status is known as "digital phenotyping" [69,92,120]. ...
Conference Paper
Full-text available
The increasing integration of computing technologies in the workplace has also seen the conceptualization and development of data-driven and algorithmic tools that aim to improve workers' wellbe-ing and performance. However, both research and practice have revealed several gaps in the effectiveness and deployment of these tools. Meanwhile, the recent advances in generative AI have highlighted the tremendous capabilities of large language models (LLMs) in processing large volumes of data in producing human-interactive natural language content. This paper explores the opportunities for LLMs in facilitating worker-centered design for Wellbeing Assessment Tools (WATs). In particular, we map features of LLMs against known challenges of WAT. We highlight how the LLMs can bridge or even widen the gaps in worker-centeric WAT. This paper aims to inspire new research directions focused on empowering workers and anticipating harms in integrating LLMs with workplace technologies.
... 2) Algorithmic profiling can be used to draw far-reaching inferences about an individual's characteristics, beliefs, desires, and behavior from online traces and other databases. Age, gender, sexual orientation, race, employment status, political opinions, preferences in food, clothes, news, entertainment, and other products, or whether one is likely to have insomnia or depression can be inferred from social media posts, search histories, traces we leave online, and data collected by personal digital devices (Huckvale et al., 2019;Loi, 2019). This algorithmic characterization is typically fluid -we move in and out of categories based on our most recent online behavior (Prey, 2018). ...
Article
Full-text available
Novel technological devices, applications, and algorithms can provide us with a vast amount of personal information about ourselves. Given that we have ethical and practical reasons to pursue self-knowledge, should we use technology to increase our self-knowledge? And which ethical issues arise from the pursuit of technologically sourced self-knowledge? In this paper, I explore these questions in relation to bioinformation technologies (health and activity trackers, DTC genetic testing, and DTC neurotechnologies) and algorithmic profiling used for recommender systems, targeted advertising, and technologically supported decision-making. First, I distinguish between impersonal, critical, and relational self-knowledge. Relational self-knowledge is a so far neglected dimension of self-knowledge which is introduced in this paper. Next, I investigate the contribution of these technologies to the three types of self-knowledge and uncover the connected ethical concerns. Technology can provide a lot of impersonal self-knowledge, but we should focus on the quality of the information which tends to be particularly insufficient for marginalized groups. In terms of critical self-knowledge, the nature of technologically sourced personal information typically impedes critical engagement. The value of relational self-knowledge speaks in favour of transparency of information technology, notably for algorithms that are involved in decision-making about individuals. Moreover, bioinformation technologies and digital profiling shape the concepts and norms that define us. We should ensure they not only serve commercial interests but our identity and self-knowledge interests.
Article
In a growing trend in digital psychiatry, algorithmic systems are used to determine correlations between data that is collected using wearable devices and self-reports of mood. They then offer recommendations for behaviour modification for improved mood. The present study consists of observations of the development of one of these systems. Descriptions of the trial emphasise the powerful role of the intrinsically motivated, responsible participant on one hand and the empowering machine learning (ML)-based technology on the other. This conceptualisation is shown to extend the neoliberal paradox of a freedom that, to be maintained, must be continually adjusted through discipline. Because of the paradoxical nature of this formulation, laboratory members disagree about the balance of agency between the objective machine learning system and the empowered participant. The guides who help participants interpret ML outputs and implement system recommendations are ascribed a replaceable role in formal accounts. Observations of this guidance practice make clear not only the important role played by guides but also how their work is relegated to the technological side of the broader formulation of the trial and further how this conceptualisation affects the way they conduct their work.
Article
Full-text available
Background: The use of social media data to predict mental health outcomes has the potential to allow for the continuous monitoring of mental health and well-being and provide timely information that can supplement traditional clinical assessments. However, it is crucial that the methodologies used to create models for this purpose are of high quality from both a mental health and machine learning perspective. Twitter has been a popular choice of social media because of the accessibility of its data, but access to big data sets is not a guarantee of robust results. Objective: This study aims to review the current methodologies used in the literature for predicting mental health outcomes from Twitter data, with a focus on the quality of the underlying mental health data and the machine learning methods used. Methods: A systematic search was performed across 6 databases, using keywords related to mental health disorders, algorithms, and social media. In total, 2759 records were screened, of which 164 (5.94%) papers were analyzed. Information about methodologies for data acquisition, preprocessing, model creation, and validation was collected, as well as information about replicability and ethical considerations. Results: The 164 studies reviewed used 119 primary data sets. There were an additional 8 data sets identified that were not described in enough detail to include, and 6.1% (10/164) of the papers did not describe their data sets at all. Of these 119 data sets, only 16 (13.4%) had access to ground truth data (ie, known characteristics) about the mental health disorders of social media users. The other 86.6% (103/119) of data sets collected data by searching keywords or phrases, which may not be representative of patterns of Twitter use for those with mental health disorders. The annotation of mental health disorders for classification labels was variable, and 57.1% (68/119) of the data sets had no ground truth or clinical input on this annotation. Despite being a common mental health disorder, anxiety received little attention. Conclusions: The sharing of high-quality ground truth data sets is crucial for the development of trustworthy algorithms that have clinical and research utility. Further collaboration across disciplines and contexts is encouraged to better understand what types of predictions will be useful in supporting the management and identification of mental health disorders. A series of recommendations for researchers in this field and for the wider research community are made, with the aim of enhancing the quality and utility of future outputs.
Article
Full-text available
We are living in an algorithmic age where mathematics and computer science are coming together in powerful new ways to influence, shape and guide our behaviour and the governance of our societies. As these algorithmic governance structures proliferate, it is vital that we ensure their effectiveness and legitimacy. That is, we need to ensure that they are an effective means for achieving a legitimate policy goal that are also procedurally fair, open and unbiased. But how can we ensure that algorithmic governance structures are both? This article shares the results of a collective intelligence workshop that addressed exactly this question. The workshop brought together a multidisciplinary group of scholars to consider (a) barriers to legitimate and effective algorithmic governance and (b) the research methods needed to address the nature and impact of specific barriers. An interactive management workshop technique was used to harness the collective intelligence of this multidisciplinary group. This method enabled participants to produce a framework and research agenda for those who are concerned about algorithmic governance. We outline this research agenda below, providing a detailed map of key research themes, questions and methods that our workshop felt ought to be pursued. This builds upon existing work on research agendas for critical algorithm studies in a unique way through the method of collective intelligence.
Article
Full-text available
The internet of things is increasingly spreading into the domain of medical and social care. Internet-enabled devices for monitoring and managing the health and well-being of users outside of traditional medical institutions have rapidly become common tools to support healthcare. Health-related internet of things (H-IoT) technologies increasingly play a key role in health management, for purposes including disease prevention, real-time tele-monitoring of patient’s functions, testing of treatments, fitness and well-being monitoring, medication dispensation, and health research data collection. H-IoT promises many benefits for health and healthcare. However, it also raises a host of ethical problems stemming from the inherent risks of Internet enabled devices, the sensitivity of health-related data, and their impact on the delivery of healthcare. This paper maps the main ethical problems that have been identified by the relevant literature and identifies key themes in the on-going debate on ethical problems concerning H-IoT.
Article
Full-text available
The conjunction of wireless computing, ubiquitous Internet access, and the miniaturisation of sensors have opened the door for technological applications that can monitor health and well-being outside of formal healthcare systems. The health-related Internet of Things (H-IoT) increasingly plays a key role in health management by providing real-time tele-monitoring of patients, testing of treatments, actuation of medical devices, and fitness and well-being monitoring. Given its numerous applications and proposed benefits, adoption by medical and social care institutions and consumers may be rapid. However, a host of ethical concerns are also raised that must be addressed. The inherent sensitivity of health-related data being generated and latent risks of Internet-enabled devices pose serious challenges. Users, already in a vulnerable position as patients, face a seemingly impossible task to retain control over their data due to the scale, scope and complexity of systems that create, aggregate, and analyse personal health data. In response, the H-IoT must be designed to be technologically robust and scientifically reliable, while also remaining ethically responsible, trustworthy, and respectful of user rights and interests. To assist developers of the H-IoT, this paper describes nine principles and nine guidelines for ethical design of H-IoT devices and data protocols.
Chapter
Full-text available
In this chapter I identify three problems affecting the plausibility of group privacy and argue in favour of their resolution. The first problem concerns the nature of the groups in question. I shall argue that groups are neither discovered nor invented, but designed by the level of abstraction (LoA) at which a specific analysis of a social system is developed. Their design is therefore justified insofar as the purpose, guiding the choice of the LoA, is justified. This should remove the objection that groups cannot have a right to privacy because groups are mere artefacts (there are no groups, only individuals) or that, even if there are groups, it is too difficult to deal with them. The second problem concerns the possibility of attributing rights to groups. I shall argue that the same logic of attribution of a right to individuals may be used to attribute a right to a group, provided one modifies the LoA and now treats the whole group itself as an individual. This should remove the objection that, even if groups exist and are manageable, they cannot be treated as holders of rights. The third problem concerns the possibility of attributing a right to privacy to groups. I shall argue that sometimes it is the group and only the group, not its members, that is correctly identified as the correct holder of a right to privacy. This should remove the objection that privacy, as a group right, is a right held not by a group as a group but rather by the group’s members severally. The solutions of the three problems supports the thesis that an interpretation of privacy in terms of a protection of the information that constitutes an individual—both in terms of a single person and in terms of a group—is better suited than other interpretations to make sense of group privacy.
Article
Even when external researchers have access to social media data, they are not privy to decisions that went into platform design—including the measurement and testing that goes into deploying new platform features, such as recommender systems, seeking to shape user behavior towards desirable ends. Finding ways to identify platform effects is thus important both for generalizing findings, as well as understanding the nature of platform usage. One approach is to find temporal data covering the introduction of a new feature; observing differences in behavior before and after allow us to estimate the effect of the change. We investigate platform effects using two such datasets, the Netflix Prize dataset and the Facebook New Orleans data, in which we observe seeming discontinuities in user behavior but that we know or suspect are the result of a change in platform design. For the Netflix Prize, we estimate user ratings changing by an average of about 3% after the change, and in Facebook New Orleans, we find that the introduction of the ‘People You May Know’ feature locally nearly doubled the average number of edges added daily, and increased by 63% the average proportion of triangles created by each new edge. Our work empirically verifies several previously expressed theoretical concerns, and gives insight into the magnitude and variety of platform effects.
Book
In this book G. A. Cohen examines the libertarian principle of self-ownership, which says that each person belongs to himself and therefore owes no service or product to anyone else. This principle is used to defend capitalist inequality, which is said to reflect each person's freedom to do as as he wishes with himself. The author argues that self-ownership cannot deliver the freedom it promises to secure, thereby undermining the idea that lovers of freedom should embrace capitalism and the inequality that comes with it. He goes on to show that the standard Marxist condemnation of exploitation implies an endorsement of self-ownership, since, in the Marxist conception, the employer steals from the worker what should belong to her, because she produced it. Thereby a deeply inegalitarian notion has penetrated what is in aspiration an egalitarian theory. Purging that notion from socialist thought, he argues, enables construction of a more consistent egalitarianism.
Book
What should be the goal of science in a democratic society? Some say, to attain the truth; others deny the possibility (or even the intelligibility) of truth‐seeking. Science, Truth, and Democracy attempts to provide a different answer. It is possible to make sense of the notion of truth, and to understand truth as correspondence to a mind‐independent world. Yet science could not hope to find the whole truth about that world. Scientific inquiry must necessarily be selective, focusing on the aspects of nature that are deemed most important. Yet how should that judgement be made? The book's answer is that the search for truth should be combined with a respect for democracy. The scientific research that should strike us as significant would address the questions singled out as most important in an informed deliberation among parties committed to each others’ well‐being. The book develops this perspective as an ideal of ‘well‐ordered science’, relating this ideal both to past efforts at science policy and to the possibility that finding the truth may not always be what we want. It concludes with a chapter on the responsibilities of scientists.
Book
A pioneering proposal for a pluralistic extension of evolutionary theory, now updated to reflect the most recent research. This new edition of the widely read Evolution in Four Dimensions has been revised to reflect the spate of new discoveries in biology since the book was first published in 2005, offering corrections, an updated bibliography, and a substantial new chapter. Eva Jablonka and Marion Lamb's pioneering argument proposes that there is more to heredity than genes. They describe four “dimensions” in heredity—four inheritance systems that play a role in evolution: genetic, epigenetic (or non-DNA cellular transmission of traits), behavioral, and symbolic (transmission through language and other forms of symbolic communication). These systems, they argue, can all provide variations on which natural selection can act. Jablonka and Lamb present a richer, more complex view of evolution than that offered by the gene-based Modern Synthesis, arguing that induced and acquired changes also play a role. Their lucid and accessible text is accompanied by artist-physician Anna Zeligowski's lively drawings, which humorously and effectively illustrate the authors' points. Each chapter ends with a dialogue in which the authors refine their arguments against the vigorous skepticism of the fictional “I.M.” (for Ipcha Mistabra—Aramaic for “the opposite conjecture”). The extensive new chapter, presented engagingly as a dialogue with I.M., updates the information on each of the four dimensions—with special attention to the epigenetic, where there has been an explosion of new research. Praise for the first edition “With courage and verve, and in a style accessible to general readers, Jablonka and Lamb lay out some of the exciting new pathways of Darwinian evolution that have been uncovered by contemporary research.” —Evelyn Fox Keller, MIT, author of Making Sense of Life: Explaining Biological Development with Models, Metaphors, and Machines “In their beautifully written and impressively argued new book, Jablonka and Lamb show that the evidence from more than fifty years of molecular, behavioral and linguistic studies forces us to reevaluate our inherited understanding of evolution.” —Oren Harman, The New Republic “It is not only an enjoyable read, replete with ideas and facts of interest but it does the most valuable thing a book can do—it makes you think and reexamine your premises and long-held conclusions.” —Adam Wilkins, BioEssays Bradford Books imprint