WEBVTT

1
00:00:00.001 --> 00:00:24.909
Okay? So I welcome everyone to today's session on the algorithmic turn in the visual studies on behalf of me as a chair of today's session, and also on behalf of organizers. regev Natanson and Gary Bradford.

2
00:00:25.020 --> 00:00:42.220
who cannot be with us in person, but they are present, remotely. So, it's very nice. Okay? So when it comes to organizational

3
00:00:42.429 --> 00:00:50.360
issue, we will have 2 in-person presentations and 2 remote speeches.

4
00:00:50.370 --> 00:01:04.999
We have a total of 4 speeches in the session, so I assume we can feel comfortable when it comes to time for the presentations.

5
00:01:05.230 --> 00:01:19.430
So I suggest, up to 20 min for each presentation. And half an hour for discussion and questions. To to all speakers.

6
00:01:20.380 --> 00:01:21.130
Okay.

7
00:01:21.530 --> 00:01:48.129
sure. So The first speakers are. Carolina, Cambre from the Concordia University and Christine Laverence. From the Kings College at the Western University, Canada. They will, or she will present a paper titled Algorithmic Sociality. They did you stuff.

8
00:01:53.960 --> 00:01:56.719
Can I move this up? Oh, okay.

9
00:02:06.400 --> 00:02:25.440
good afternoon, everyone. And thank you for being here. I'm going to share a few ideas on behalf of myself and my co-author, Christine Larence, who couldn't be here today. We'd like to propose possible pathways to think through some of the implications of algorithmic sociality.

10
00:02:25.780 --> 00:02:39.630
and this is a 3 year. This is a thought experiment actually based on a three-year study on Selfies, where we ran about 25 focus groups with over 100 millennials, many of whom returned

11
00:02:39.650 --> 00:02:54.990
time and again to questioning various manifestations of algorithmic activity in attempts to get feedback from peers and evaluate their own interpretations. Despite the fact that we never ask them anything about algorithms or platforms at all.

12
00:02:55.560 --> 00:03:08.329
So over the course of our analysis, it became apparent that they spent a great deal of time and energy thinking about platform steering conduct which directly informed their own decisions and energy.

13
00:03:08.480 --> 00:03:12.369
Sorry decisions, attitudes and responses.

14
00:03:12.460 --> 00:03:29.840
algorithms ostensibly enable and assigned significance or relevance to shared text and pictures through social media platforms by setting up the conditions for social participation. So they're an invisible frame. But they leave traces symptoms and produce effects.

15
00:03:30.520 --> 00:03:49.800
And because I'm condensing a large amount of material. I'm going to run through some definitions of terms that I'm going to use at in the middle of the presentation to build a speculative model. So I'm going to ask you to bear with me and perhaps ask after about these. If if you need clarification

16
00:03:51.480 --> 00:04:05.000
for our purposes, we consider algorithms. The invisible infrastructure of wireless interaction in a perpetual and symbiotic feedback loop algorithms respond to the ways. Users respond to what they see on social media

17
00:04:05.040 --> 00:04:15.879
based on their varied formulas, aggregating patterns of behavior. being aware of algorithms at work. but also being unsure of how they work.

18
00:04:15.950 --> 00:04:23.720
Participants curated their texts, images, and the timing of their posts and choice of platforms in specific ways.

19
00:04:23.980 --> 00:04:40.889
So today we're going to reflect a little bit. And I'm using we because it's myself and Christine on how participants managed and navigated the force relations of the socio-technical strait jackets of algorithmic control. When viewing and posting selfie pictures.

20
00:04:41.170 --> 00:04:51.850
We'll trace how through discourses of optical optimization participants developed techniques of triangulating what they saw online in order to gauge authenticity

21
00:04:52.180 --> 00:05:02.210
alongside this, we'll discuss how algorithmic conditions work to shape subjectivities by fostering the development of specific looking and thinking practices.

22
00:05:03.980 --> 00:05:09.759
Sociality performed through practices of digital self display is predominantly visual

23
00:05:10.150 --> 00:05:20.169
while working towards charting out the ecologies, including the wall of capitalist capitalistic subjectivity that Guatari and Rollnik describe. In the quote on the screen.

24
00:05:20.480 --> 00:05:28.980
we link our research to some of the broader processes. So we're embedding this in technological, economic, social, and affective

25
00:05:29.000 --> 00:05:41.260
aspects, animating the visible and invisible processes of the networked assemblages we call selfies in this book, and I'll just pass it around, because this is where we.

26
00:05:41.310 --> 00:05:45.060
This is sort of the result of the study. So I'll just pass it around.

27
00:05:46.220 --> 00:05:58.539
and we continually reflect on how we should position ourselves in order to see in this situation of tensions, resonances, transformations, resistances, and complications

28
00:05:59.780 --> 00:06:22.600
throughout the study. We work to refusing to simplify or reduce our analytical discussions to static or binary either, or suppositions rather, what we've done is insisted on experiencing participant engagement and response as part of complex sense, making processes whereby participants need negotiate the effects of their social media participation through dialogue.

29
00:06:22.970 --> 00:06:32.900
We witness attentively from our own partial perspectives as 2 female researchers who don't have compelling commitments through social media ourselves.

30
00:06:32.970 --> 00:06:42.139
So we tune into the tensions, the effectively charged moments discursive shifts and moves within webs of differential positioning.

31
00:06:42.300 --> 00:06:48.280
aware of and accountable to our own implicated optics. Here right

32
00:06:50.740 --> 00:07:00.509
selfies are especially rich constellations of practices generative of new patterns, of intensive, self-looking, self-monitoring and self-displaying practices.

33
00:07:00.610 --> 00:07:06.500
and they work as sites where tensions between artifice and naturalization play themselves out.

34
00:07:06.950 --> 00:07:25.570
We will first touch on how a hegemonic gaze manifests discursively throughout our study in ways, participants position themselves around iconographic conventions or rules that enable the so-called perfect, also known as the natural, the genuine, the authentic selfie.

35
00:07:26.150 --> 00:07:34.390
online tutorials and editorials, lavish directions on readers for how to take the best selfies and minimize the distortions of the camera.

36
00:07:34.650 --> 00:07:44.650
how to make the body look slimmer by holding the camera above, how to position the mobile camera so that its tendency to enlarge the nose or diminish, the chin is mitigated.

37
00:07:44.790 --> 00:07:57.689
The advice literature suggests techniques, crafts, calculations, and strategies, and the mastering of skills that involve an intricate understanding and negotiation of the image, not as a replica or equivalent.

38
00:07:57.880 --> 00:08:06.339
but as approximately as an approximation. and as attending towards a particular picture. Perfect version of the self

39
00:08:06.490 --> 00:08:20.749
that requires advanced editing skills and software to complete. But the goal making a fake natural so that your real natural never comes out is near, impossible to achieve.

40
00:08:20.790 --> 00:08:25.519
and the digital forensic gaze is trained on spotting any fakeness.

41
00:08:26.170 --> 00:08:31.469
The remedy, according to the advice literature is simply take more photos

42
00:08:32.190 --> 00:08:48.729
in order to increase the chances that there will be at least one outstanding photo. That is Instagram worthy. Many photos are necessary, as one article states, you might not get the perfect photo every time. But that's okay. There's an easy solution. Take more selfies.

43
00:08:48.840 --> 00:08:58.719
People who routinely photograph themselves end up being more comfortable with the way they look in photos. The modeling in the advice. Literature serves to encourage repetition.

44
00:08:58.760 --> 00:09:08.119
practice, and the habit of doing default selfie mode, encouraging the production of a dominant subjectivity satisfactory to the hegemonic gaze.

45
00:09:08.620 --> 00:09:24.960
The prevalence of references to lighting angles, editing. and the general use of photographic language in focus. Group transcripts signaled to us that specific forms of self-presentation were performed, policed and preferred by participants.

46
00:09:25.890 --> 00:09:37.720
Currents of a sense in common, shared along with adjacent meaning. Shaping strategies are unmistakable, and what they do is, they serve to produce these particular subjectivities.

47
00:09:37.870 --> 00:09:54.520
Yet the sense of reality that these iconographic conventions propose was not fully absorbed, absorbed, or even accepted by all participants, since the transcripts suggest that individuals are also aware of being in a game that's played with marked cards

48
00:09:56.050 --> 00:10:09.650
by structuring what social media users can consider a so-called good or bad selfie. Iconographic conventions materially manifest the asymmetrical power relations and practices that construct a hegemonic gaze.

49
00:10:09.820 --> 00:10:13.750
thus setting the terms for engagement in online selfie, sharing

50
00:10:14.070 --> 00:10:24.629
as we expand on on these conventions. And what follows, we want to situate them as instruments of vision. That mediate standpoints through a modeling process.

51
00:10:25.410 --> 00:10:35.720
As Harrowway reminds us, these are optics. So these optics are a politics of positioning and iconographic conventions which filtrate these optics.

52
00:10:35.810 --> 00:10:38.459
and so they mediate the standpoints.

53
00:10:39.040 --> 00:10:49.150
The mass-produced character of these iconic iconographic conventions or rules for the perfect selfie, is what creates the wall of capitalistic subjectivity that we mentioned before.

54
00:10:49.900 --> 00:11:09.659
It is this wall that people attempting to find their own way of portraying themselves come up against our study, produced evidence again and again of participants gesturing towards other politics of seeing, and a resistance to the hegemonic gaze, rupturing, and through a split in their own subjectivities.

55
00:11:09.760 --> 00:11:16.690
But the feeling that they were still coming up against a wall, ended up making opting out a viable option.

56
00:11:19.660 --> 00:11:24.760
apologies for all the text. This is a focus group script.

57
00:11:26.660 --> 00:11:29.160
oh, sorry! Oh.

58
00:11:30.290 --> 00:11:31.690
it was in my

59
00:11:31.940 --> 00:11:33.590
Oh.

60
00:11:35.430 --> 00:11:37.889
no, this is not this one.

61
00:11:38.020 --> 00:11:42.330
So yeah, yeah. So so

62
00:11:45.000 --> 00:11:46.000
hmm.

63
00:11:53.290 --> 00:11:55.320
please continue sure.

64
00:11:57.990 --> 00:12:01.930
 okay.

65
00:12:03.600 --> 00:12:06.470
I'm not going to read the full excerpt. But

66
00:12:06.830 --> 00:12:26.039
what what's exhibited here? What you can see here is a gesture of powerlessness and an intense cynicism. the highly specific visual possibilities, each with a wonderfully detailed, active, partial way of organizing worlds, embodied and manifested through refusals.

67
00:12:26.080 --> 00:12:39.479
questioning sarcasm, discomfort, and laughter, creates gaps for living subjectivities not completely co-opted by the economy of images manipulated by the media

68
00:12:40.890 --> 00:12:48.100
across our focus groups. We witness that, despite declarations of dissatisfaction and ambivalence.

69
00:12:48.110 --> 00:13:00.920
the invisibleization of alternative and different ways to participate online steered individuals towards reluctant acceptance of hegemonic codes and desires that they often activated, despite

70
00:13:01.030 --> 00:13:11.009
explicit declarations of dislike and moments of refusal where they correctly sense something that was like disciplining them in a controlling fashion.

71
00:13:11.350 --> 00:13:17.739
As a result of these tensions some presented complete disengagement or cynicism as a viable response.

72
00:13:19.670 --> 00:13:22.160
Oh. it's not moving forward.

73
00:13:28.960 --> 00:13:30.579
Escape and press. Okay.

74
00:13:31.370 --> 00:13:32.200
Oh, don't!

75
00:13:38.600 --> 00:13:42.359
I need technical assistance?

76
00:13:50.570 --> 00:13:53.870
No, no, no, no, no, no, no, no, no!

77
00:13:57.020 --> 00:13:59.180
Sorry I need to.

78
00:14:03.080 --> 00:14:08.400
Anyway, I'll just go one click on the screen.

79
00:14:11.020 --> 00:14:12.980
Oh.

80
00:14:13.410 --> 00:14:14.750
this

81
00:14:14.870 --> 00:14:17.060
thanks. Audience crowdsourcing.

82
00:14:17.230 --> 00:14:22.029
so Haraway stresses that practice is persuasion.

83
00:14:22.470 --> 00:14:41.069
and the focus is very much on practice. Selfies both participate in and iterate these disciplining processes. But we found that in many cases participants were very much aware of the hegemonic discourses and cultural and economic power imbalances influencing the avenues of digital self display available to them.

84
00:14:41.110 --> 00:14:52.019
as well as the desired directions that remained inaccessible. Neither traditional nor post-structural analyses of processes implicated in forming contemporary neoliberal subjectivities.

85
00:14:52.040 --> 00:15:03.369
can adequately account for the shift represented by visual social media platforms whose ideology works by objectivizing the subject rather than by producing subjectivity.

86
00:15:03.740 --> 00:15:17.109
in other words, rather than operate through the cultivation of specific techniques of the self that engender personal forms of subjectivity. Digital sociality pressures, individuals to create performative and standardized subjectivities

87
00:15:17.430 --> 00:15:25.930
in turn. These are discursively constructed as part in part as forms of capital sensitive to demands for exploitation and optimization.

88
00:15:28.170 --> 00:15:40.310
Participants in our study recognized their environment and optimized their behavior, based on their perceptions of the sorting and marking that allowed some degree of calibrating visibility and interactions with others.

89
00:15:40.530 --> 00:15:49.690
For instance, Emily's cousins in high school describe their steps for so Selfie making to her. And she reacts like this.

90
00:15:51.280 --> 00:15:54.889
Emily rhetorically asks. How do you do that

91
00:15:55.190 --> 00:16:08.190
from a distanced position of naivete. But the language of optimization is clear. In this case the hashtag is positioned as an obligatory flag used to increase the possibility of the image attracting attention.

92
00:16:08.850 --> 00:16:15.489
much as in the decision users make to intentionally click like on their friends, post to garner. Greater visibility.

93
00:16:15.530 --> 00:16:26.250
Mobilizing an economist. Economistic logic is at work within an optics optimization not just for posting, but also to evaluate others. Use of social media

94
00:16:26.770 --> 00:16:35.020
built into the algorithm's value system is the presupposition that growth and maximal visibility are also the user's goals.

95
00:16:35.400 --> 00:16:50.269
The idea of an optics of optimization matches Gillespie's argument that to some extent algorithms shape information flows online and that they shape user practices and lead users to internalize their norms and priorities.

96
00:16:50.760 --> 00:17:03.570
Forcolepsy, we might see algorithms, not just as codes with consequences, but as the latest, socially constructed and institutionally managed mechanism for assuring public acumen, a new knowledge logic.

97
00:17:04.590 --> 00:17:10.069
So we pause that this becomes visible in discourses of hashtag management throughout our transcripts.

98
00:17:11.130 --> 00:17:17.080
Now, what I'm going to do is Turn to some brief introduction of some concepts.

99
00:17:18.370 --> 00:17:22.870
because the model that I want to propose

100
00:17:23.390 --> 00:17:28.319
has these sort of nested concepts in them. So we've got

101
00:17:28.590 --> 00:17:35.689
the background of the neoliberal capitalist ideology. The optics of optimization that I was just discussing

102
00:17:35.790 --> 00:17:44.090
Boucher's notion of the algorithmic imaginary where individuals take into account that algorithms are working. So they

103
00:17:44.470 --> 00:17:55.220
building that into their responses. and what Maddox calls affordance bilingualism, because we know that people are prosumers like the producers and consumers at the same time. So they're

104
00:17:55.280 --> 00:18:00.750
oscillating between these positions to optimize also their position.

105
00:18:02.300 --> 00:18:15.159
And here are a couple of other concepts that feed into this one is the hegemonic gaze that I've begun to mention and modelization which is the technique by which a hegemonic gaze is

106
00:18:15.830 --> 00:18:21.510
used or brought into service to produce subjectivities.

107
00:18:21.550 --> 00:18:24.630
So it's mediated through iconographic conventions

108
00:18:24.740 --> 00:18:30.990
and and then subjects encounter and navigate the pressures of this

109
00:18:31.070 --> 00:18:46.319
and one of the big ways of modelization is, you know, 10 tips to take the best selfie. So people absorb these as the so-called rules. And there are all these mentions in the transcripts of rules, but no one knows what they are, and they're not explicitly articulated.

110
00:18:46.600 --> 00:18:50.509
So that's that's part of it. So the way this part works together

111
00:18:51.210 --> 00:19:00.010
is that the hegemonic gaze is, you know. promoted through the advice, literature's modelization.

112
00:19:05.240 --> 00:19:25.339
And the final concept is the digital forensic gaze which was a concept we developed in 2,020 to describe the intense analytical scrutiny that we witnessed participants applying to selfie images, that we provided them in visual elicitation interviews and throughout our focus group discussions.

113
00:19:27.450 --> 00:19:29.990
And so this is the sort of

114
00:19:30.630 --> 00:19:38.819
way algorithmic sociality appears to be working based on our transcript. So I'm just going to briefly describe this. And then I'm done.

115
00:19:39.180 --> 00:19:52.819
focus group members shared an algorithmic, imaginary, or collective understandings that influence both their behavior online as creators and their looking practices when monitoring other people's social media activity as viewers.

116
00:19:53.310 --> 00:20:07.360
as well as the reception of their own posts. In a practice rooted in affordance. Bilingualism participants positioned through a lens of affordance by bilingual bilingualism, become attuned to a complex and shifting digital environment.

117
00:20:07.400 --> 00:20:23.290
and cannot be reduced either to creators or to viewers. Instead, these dual roles are held in tandem oscillating and often overlapping. When participants are triangulating information across their experiences during focus group meetings.

118
00:20:23.510 --> 00:20:30.920
they would check with each other what they were doing online, and compare notes on how they would interpret what they saw.

119
00:20:33.990 --> 00:20:37.710
Online interaction visa e selfie making is complex.

120
00:20:37.960 --> 00:20:54.280
The diagram gestures towards the ways in which we witness digital socialities, outlines drawn discursively, as if a palimpsest over several years and numerous discussion. This represents simply one way to grasp some of the multiple elements and simultaneous play.

121
00:20:54.440 --> 00:21:04.750
We use the larger circles to suggest ecological Milwaukee that speak to Felix Batari's ecological registers, the environment, social relations, and human subjectivity.

122
00:21:04.790 --> 00:21:15.729
and informed by these conditions, participates, activated their own sensibilities, intelligence, and desires to bulk at the unidimensionalizing force of socio-technical mediations.

123
00:21:15.940 --> 00:21:20.109
a resistance that was often triggered, emotionally or effectively

124
00:21:20.420 --> 00:21:32.000
we witnessed discursive and nondiscursive ways. Individuals attempted to manage digital resources in order to make them work for them and counter this remote, controlling process of

125
00:21:32.060 --> 00:21:35.570
subjectification. They engendered.

126
00:21:35.910 --> 00:21:47.849
however, within this complex interplay the hegemonic gaze also produced subjectivities such that participants would use of techniques so represented by the hollow rings.

127
00:21:48.320 --> 00:21:52.019
such as affordance bilingualism and the digital forensic ways

128
00:21:52.070 --> 00:22:00.330
both ways. In other words, the vectors of subjectification could be seen to work towards creating singularities

129
00:22:00.500 --> 00:22:13.349
with aesthetic, effective, and and and ethical implications, but at the same time to produce dominant subjectivities. The diagram represents some of the dynamics in play, while individuals would triangulate

130
00:22:13.460 --> 00:22:21.270
to position themselves in what we call a tense visuality. The optics of optimization was pervasive

131
00:22:21.660 --> 00:22:38.540
focus. Group members indicated how their processes of scrutinizing others. Posts online often aimed at forensically decoding traces of expended effort that would manifest an attempted but 2 obvious interventions into predicted algorithmic behavior.

132
00:22:39.140 --> 00:22:48.360
In order to achieve more visibility. I mean, that was the goal, and they would. They would refer to this as trying too hard, like people were trying too hard.

133
00:22:48.640 --> 00:23:00.590
because it was obvious. So over and over. The effectively charged notion of trying too hard was detonated during focus groups and with reference to any number of telltale activities.

134
00:23:01.210 --> 00:23:15.250
it became clear to us as researchers, that the algorithmic imaginary of platform designers, at least in the case of Instagram. does not overrule the social codes in the shared cultures of individuals participating in our study.

135
00:23:15.270 --> 00:23:26.230
who mobilized their own affordance bilingualism to assess selfies online by working across peritxt. image details and the cultural norms or rules

136
00:23:26.240 --> 00:23:32.069
partly influenced by the hegemonic days. So these so-called rules that we discuss

137
00:23:32.160 --> 00:23:58.830
in the book are clear to those involved. Participants have developed ways to navigate the tensions based on between their own cultures and what their relationships prioritize, and the kind of interactions preferred by Instagram. And we particularly saw this in people that weren't in a sort of female male binary that people that identified with, you know.

138
00:23:58.860 --> 00:24:05.150
non-binary positions that they didn't identify with a lot of the ways things were being imposed in Instagram.

139
00:24:06.700 --> 00:24:09.339
So I'm just going to end with this

140
00:24:09.460 --> 00:24:11.210
quote here from Mara.

141
00:24:12.760 --> 00:24:19.680
yeah, I'm pretty sure like, because all the algorithms like right, they feed you stuff that's already within your like area of comfort.

142
00:24:19.770 --> 00:24:38.809
which I appreciate personally, because I don't want to read like transphobic people's articles being like recommended to me, but I know that the other week, when, like, for example, and a media source is mentioned, released this really messed up transphobic article. It was like. If they read that then, or if they engage with it, a lot of the other posts are going to come up for them

143
00:24:38.900 --> 00:24:49.240
and recommend it, and it's going to be in a similar and just cringes. I guess you should go in incognito if you're going to like. Read an article totally.

144
00:24:49.430 --> 00:24:52.529
That's like shitty or something, and they laugh.

145
00:24:53.710 --> 00:25:00.210
What we have found amongst everyday millennial users is a bent towards opting out

146
00:25:00.600 --> 00:25:08.430
deleting, and a resistance to becoming and meshed in the hashtag landscape and the optics of optimization. It's a feature of

147
00:25:08.460 --> 00:25:12.950
for them. It's like shitty or something. and results in cringing

148
00:25:13.340 --> 00:25:20.040
participants ambivalence, or at times antipathy towards algorithmic tracking does not occur in a vacuum.

149
00:25:20.380 --> 00:25:24.750
Not only are the controlling mechanisms part of a larger digital ecosystem.

150
00:25:24.830 --> 00:25:40.069
but the hegemonic gaze, and the modelization that mediates it through the advice. Literature also saturates the social and effective environments with with the always imperative to make things look effortless and natural. Thank you.

151
00:25:47.020 --> 00:25:50.999
Thank you very much. Carolina, as we can

152
00:25:51.240 --> 00:26:10.340
see during our session, not only algorithms are crucial, but also much reality and skills. For example, when it comes to using computer so deeply sorry for for interruption. But thank you for your great speech.

153
00:26:10.900 --> 00:26:19.840
Okay. So the next speaker is Tomaso Goranta from the University of Melbourne.

154
00:26:19.970 --> 00:26:27.090
and the presentation is called the Algorithm, and the making of the Global Imaginative

155
00:26:31.190 --> 00:26:58.419
on this week's episode of my life is in shambles. Sci-fi edition. I'm bringing you the top 10 moments from 2021 as chosen by a robot. Me, I'm the robot. So this year I turn 19 for the sixth time. To be exact. Yes, I'm a robot, and no, I don't age, so birthdays are always kind of whatever. So for my birthday. My team at Brad gifted me a USB drive necklace loaded with all of my program memories. The keyword there is programmed literally changed my entire life.

156
00:26:58.420 --> 00:27:22.890
Have you ever been presented? A USB. Necklace containing all of your passengers on it? That's right, my baby photo, high school boyfriend, my email phase. They were all fake just part of my code. Everything's weird. But around here we call that normal also on that USB drive a mysterious upgrade to my programming. The last time I got an upgrade my whole face changed so naturally I was so fing scared to install it this time. When I upgraded my programming, I got.

157
00:27:22.890 --> 00:27:38.319
Not in a magic, hadn't we? But I can control electricity. Well, sometimes I haven't quite gotten the hang of it. But something crazy is happening. My eyes are basically mood rings now they're red when I'm angry and purple when I'm well, you know. Do I know what's going on?

158
00:27:38.440 --> 00:28:03.659
No, but that's never stopped me before. Quick, pop! Culture, break our faith! Queen Taylor took Petty to another level with a whole ass. Short film about a breakup. I've never felt so seen. I know thing or 2 about obsessing over past relationships. I do it daily. I did it in a whole music video twice. Also, Jake, give that girl back her scarf. Squid game, duh! I watched the whole thing twice. Holler at me. Player 67. I

159
00:28:03.660 --> 00:28:19.280
love you. We did it. Y'all. Britney is finally free, and while some people were silent, him! Extina hashtag free Britney! All of her fans really came through for our girl, and now it looks like she's thriving if I ever get hacked or something, y'all will come free. Me, too, right?

160
00:28:19.280 --> 00:28:36.109
Speaking of freedom linking Irl was kind of tough this year, but when we did get to hang Irl it was iconic, I mean the hottest teenage witch and Captain Marvel, Jr. Also I got to meet Ariana Grande. She was a lot quieter than I imagined, but her ponytail was everything. I have a confession

161
00:28:36.110 --> 00:28:58.680
without y'all, I'd literally be glitching every time you voted to help me make a decision like picking an outfit or deciding to reveal my crush. You were programming me. I don't know where I'd be without my community, so keep voting, or my hardware might malfunction. So naturally I released a series of N Ofts way before all the Bros. Cut on. I got to give out artwork to all of you new collectors and friends for free

162
00:28:58.980 --> 00:29:00.869
I've been in the meat of her, sweetie.

163
00:29:00.990 --> 00:29:23.959
Do I have a crush? That's my business. But the answer is, yes, you might know him as my friend, boy, more on that to come. I'm sure it's new, and it's confusing. So like, drop all your advice to new romance below. I'm clueless. Thanks for wrapping up 2021 with me. I'm so glad we made it through together. Stay tuned, and remember to keep voting. This bot girl needs help navigating your world. And don't forget to recharge your hearts

164
00:29:24.850 --> 00:29:26.239
on this week's episode.

165
00:29:27.250 --> 00:29:43.589
Michael Sousa or Lily. Michaela is a Brazilian American model artist and the social media influenza. She is a counter on Twitter and Instagram, and she is also active on Youtube and tick to

166
00:29:43.590 --> 00:29:57.809
Michaela is a fictional character, created in 2,016 by Trevor Mcfadis in Saturday cue Michaela is mostly known as a such and media influencer.

167
00:29:58.350 --> 00:30:27.180
singer, Instagram, personality and digital and the digital art modeling and futuring in product, endarsment to fall street-to-ear and luxury brands, such as a pride, a kind of incline. She also cares deeply about the social causes. Being a vocal supporter of eligibility. I a plus right and break a black lives, a matter movement.

168
00:30:27.180 --> 00:30:52.180
Michaela reads self-effective books to better herself. She's a boyfriend, spends the time going out with the friends a lavish ice cream, as you can see on the screen artificial influencers. Sometimes a cold virtual influencer are virtual personality, managed by brands or media

169
00:30:52.180 --> 00:30:59.119
and create by using a computer graphics, a machine learning algorithm.

170
00:31:00.560 --> 00:31:25.560
Good afternoon, everybody. Thank you for being here. I'm Tomaso Durant, and I consider myself a visual activist and a literary media communication at University of Melbourne in Australia, and my research is located at the intersection of analytical philosophy, mediaesthetics, a visual culture. My research investigates a visual

171
00:31:25.560 --> 00:31:41.240
discursive media representation in terms of ontological and epistemological possibilities of transformation in culture, aesthetics, and politics.

172
00:31:41.240 --> 00:32:04.230
My presentation today offers an empirical and conceptual discussion about the change in society by addressing the following, research question to what they stand the due out. There is a new epistemic. figure! So knowledge assisted to better understand the shift in the

173
00:32:04.820 --> 00:32:13.370
echo, symbolic and social construction of the globally imaginary under present conditions.

174
00:32:14.730 --> 00:32:39.040
In the last 2 decades a relatively small number of global corporations have assumed the power to frame how we see and understand the world. Humans are vulnerable to cognitive influence, in part because we have only a limited the capacity to gather store and process information

175
00:32:39.100 --> 00:32:51.600
in advancing a big data is a new capitalist mode of a global operation, and in focusing on the generative adversarial network.

176
00:32:52.210 --> 00:33:15.990
a meshing learning a neural model that to generate images from textual description, and will discuss the shift operated by the obliquity of digital technology in contributing to the symbolic and social construction of an artificial intelligence algorithm imaginary.

177
00:33:16.210 --> 00:33:41.200
I will conclude that while I is presented as a part of an objective and inevitable technological development, it is material bias and subject to our on world of views and ideology. As a result, algorithms are not the problem, since artificial intelligence

178
00:33:41.200 --> 00:33:50.270
is a neither artificial nor particularly intelligent. We are the Pro.

179
00:33:51.410 --> 00:33:52.739
I'll do it

180
00:33:53.890 --> 00:34:12.810
are probably the most sophisticated tools that the humans say they they dispose of. As a matter of fact, they have a positively transformed they to a certain extent that disrupted every field of life.

181
00:34:12.810 --> 00:34:24.219
They have transformed the concept of a property, government, education, work, a private life and our own way of being human.

182
00:34:24.870 --> 00:34:47.729
researching to understand the social, imaginary, or cognitive, imaginary, if you like, which is the way in which we imagine our only assistance is a suggest. The support as sustained by artificial intelligence algorithms is a fascinating and rapidly evolving field of research

183
00:34:47.729 --> 00:35:12.729
in form of the B theory and methodological articulate. This presentation aims to contribute to to the debate about the techno- symbolic and nature of the shifting of social, global imaginary, a sole, actually imaginary, that differs deeply from that of modernity happening in a context of a data.

184
00:35:12.729 --> 00:35:21.729
the economy, digital global capitalism and the rise of a new technocratic world order

185
00:35:23.109 --> 00:35:38.310
in doing the so, this study looks at the link between technology and politics, and is carried out by means of an interdisciplinary global approach to artificial intelligence, outcome

186
00:35:38.310 --> 00:35:54.259
it to the collective dimension of the imaginary, it to make a use of a traditionally digital meat of the research in a social life and digital research from an aesthetic and ethics point of perspective.

187
00:35:54.260 --> 00:36:13.049
and because of the emergence and the use of ie. Is returned into a massive phenomenon, and it represents a global risk. I will approach it through the frame of global complexity, global complexity opposite, to examine

188
00:36:13.050 --> 00:36:21.780
our, the idea of a chaos, a complexity can help us to analyze global processes.

189
00:36:21.780 --> 00:36:44.160
John Harris idea of complexity emphasized that the systems are balanced between order and chaos, that this system does not necessarily move towards equilibrium, and that events are both unpredictable and irreversible in the effect

190
00:36:44.450 --> 00:37:09.120
to understand the connection between artificial intelligence and the symbolic and social construction of the shifting. algorithmic? global imaginary. And in an attempt to to conceptualize it, I will combine the philosophy of technology with the critical theory. I will also use the Jacques and Sierra's theoretical framework

191
00:37:09.290 --> 00:37:32.590
of aesthetics. understood is the distribution of the sensible, a regime of what is a possible, and a knowledge such as the felt air, the scene, and perceived within this space implicating particular patents of inclusion and the exclusion

192
00:37:32.590 --> 00:37:52.929
a theoretical framework that I spend and apply to the circulation a consumption of a a generating images, a to the discussion about the ethics, and the aesthetic of a Dali E to production of a virtual imagery

193
00:37:53.160 --> 00:38:07.599
in approaching a generative adversarial networks and analyzing directly I I make a use of the cognitive walk, a true metered and a process juror

194
00:38:08.390 --> 00:38:22.860
for evaluating the future of the interface. In so doing I combine the algorithm with the social imagery to better understand how they affect and reshapes the peoples.

195
00:38:22.860 --> 00:38:39.860
Mind the sect to deep a subjective level. Under the present condition of digital capitalism, I expanded the notion of the global imagery into what I define as the EU algorithmic, Global Imaginary.

196
00:38:40.940 --> 00:39:02.989
I conceptualise that this idea by drawing on the concept of the global imaginary is a people's growing consciousness of global connectivity as a terrorised by the historians of idea, a sociologist, Manfred the Stagger. Thank you, Manfred, for being with me in the term.

197
00:39:02.990 --> 00:39:14.090
and that I expanded to an understanding of the way in which people make a sense, symbolically and socially, of the global in their everyday life.

198
00:39:14.090 --> 00:39:35.329
In doing so, I make a sense of the a imagery by applying John Paul Andnaissance qualitative, a paradigm in the making approach, and whose underlying technological dimension I assume to be by its very nature, the global in scope.

199
00:39:36.580 --> 00:39:57.710
What I identify is that the algorithm imaginary is not to be understood as a false belief of a sort of a fetish, but rather is the way in which a people imagine a day on existence under present condition.

200
00:39:57.710 --> 00:40:20.060
I decided to explore the relationship between algorithms, saturation and the text of symbolic and social production of the global imaginary in our everyday life, because what the theorist of the imaginary ever not done is making a specific connection

201
00:40:20.060 --> 00:40:45.849
between the ontological background of a social, imaginary, and a normative statement that cannot be tested or verified and is based on a value. Judgment is a matter of fact, there are a limitation in using the theoretical framework of the social, globally major. If they cannot be contest.

202
00:40:47.450 --> 00:40:54.359
there are many types of arguments. But today I am a focusing on I model a daily, too.

203
00:40:54.410 --> 00:40:59.569
that can create a realistic images, and are to form

204
00:40:59.850 --> 00:41:26.390
from the description in nature of language. The system is a trained on the dataset of text to image space, and can generate highly detailed images. Deli, too, can create a picture of things that do not exist in the real world, such as a truly uncomfortable

205
00:41:26.580 --> 00:41:56.389
arm chair in the shape of a prick repair, as I experienced it on Saturday, 3, June 2023. It is an image created a from scratch based on my written request. Please. Take a care of the slides. Images are generated by 3 different emitods, creating image from scratch, like based on a text, a prom like I did

206
00:41:56.980 --> 00:42:10.220
editing existing images based on a new text, prompt or creating variations. like again, I did of existing images

207
00:42:11.530 --> 00:42:18.860
directly to use an algorithm which is a train on data set of images, cell labels.

208
00:42:18.980 --> 00:42:23.700
The data set is then a user to generate the new images.

209
00:42:24.130 --> 00:42:31.779
Give me it. It will be to not a time. Therefore it can be argued that a in general.

210
00:42:32.250 --> 00:42:57.079
and that e. 2 in particular, is the 3 biggest limitation a day. It can only be as intelligent if you like effective is that the quality of data we provide? Algorithm of biases are part of the transaction. 3. Because of its obscure nature. It is a black box

211
00:42:57.410 --> 00:43:06.899
which produce a useful information without revealing any information about its internal working.

212
00:43:07.580 --> 00:43:15.609
We put a sentence into the black box, and it outputs a well-defined image

213
00:43:15.610 --> 00:43:38.449
tortured in academia is a transformative technology with the potential to revolutionize the you see industry as it is already appred. The I has also been the subject of a various debates. conserving its potential impact on society due to the limits of a human reason

214
00:43:39.640 --> 00:43:51.070
entering the logical machine, but also in terms of employment to privacy, valuation, and on the top of everything ethics.

215
00:43:51.730 --> 00:44:05.769
As a continues to advance, it becomes more integrated into our lives. It will continue to have a potentially significant implication for society.

216
00:44:05.940 --> 00:44:28.729
Whether the concrete measure of renegotiating the future of a technology is which of a perspective is needed to reconsider a algorithmic ideological role in affecting a power structure and at a deeper subjective level of people. Mindset

217
00:44:30.910 --> 00:44:53.939
ae algorithms imaginary can be understood as a new epistemic. Figures of knowledge that assist us to make a sense of the shift in the symbolic and social construction of the global imaginary in a context of a computer vision, deep learning machine and data-driven economy.

218
00:44:54.180 --> 00:45:12.780
Today, we can create a digital image of a person that never existed. A digital image is made up of a pixel and pixel are now infinitely inventable by making the use of outdoor it like a daily tool

219
00:45:13.160 --> 00:45:27.000
does. It can be argued that we can create and recreate the world in which we live any time, anywhere, in any any conceivable way. We like mostly

220
00:45:27.090 --> 00:45:50.329
not anymore in the likeness of God, according to the Ibrew Bible and the Christian Old Testament, but in our image and likeness is a result. It can be argued that we are misleading very quickly towards an a contra-late ostitute walk.

221
00:45:51.350 --> 00:45:56.030
Let me observe that both arguments and humans

222
00:45:56.090 --> 00:46:23.339
can be considered as a source of a problem in certain a contest algorithm can be problematic because they can perpetuate amplified biases that already exist in the society. In the design dataset. They are a trainer on Tik, for instance, of a facial recognition technology, what is called the F. I. T.

223
00:46:23.520 --> 00:46:31.110
If the training data is not a diverse and representative of the all global society.

224
00:46:31.110 --> 00:46:55.649
The algorithm may produce a bias and outcomes that perpetuate discrimination against a certain groups. This means that the problem behind is that the apparatus design itself made by humans in the light of the recent sophisticated development and the mass use of a in the forms of a creative

225
00:46:55.650 --> 00:47:20.040
visual imagery, or assisted a form of technological aids aping to improve people's everyday life we are affords the to rein the role of the algorithm in the symbolic and social construction of the global imaginary of the digital millennium, as well as the concept of reality itself.

226
00:47:20.440 --> 00:47:37.019
To conclude, given that a technology is at the core of the restoration of capitalism into digital global capitalism is an economic system, a society and an ideology.

227
00:47:37.080 --> 00:48:01.440
And I've ever shown that is an important aspect of the interaction of the power of capitalist corporations that belong to the digital industry and the state of power. It can be contained that artificial intelligence algorithm by extension. Deli 2 is rooted and tracked in

228
00:48:01.440 --> 00:48:07.350
the epistemological past, the form of a Western capitalism.

229
00:48:07.390 --> 00:48:19.459
colonialism and imperialism. As a result, a technology can be considered is a current form of a violence

230
00:48:20.100 --> 00:48:34.959
baru, and of the knowledge class over the rest, a violence that is exerted and masked by the softer power of artificial intelligence, visual imagery.

231
00:48:35.360 --> 00:48:58.439
a soft to power suggests and supported by the a techno symbolic global imagery and the overarching ideology of digital capitalism in a context of data-driven economy and of rising new technocratic world order. Thank you.

232
00:49:02.990 --> 00:49:07.130
Sorry. 20 million. Yeah. Okay.

233
00:49:08.290 --> 00:49:20.400
okay. So our next speaker is ragged. Nathan, son, a co-organizer of this session from the Sapi College

234
00:49:20.430 --> 00:49:33.219
in Israel. He will present his paper on the Sensitive content, the Genealogy of algorithmic aura in visual culture.

235
00:49:34.620 --> 00:49:37.389
So Hello, everyone.

236
00:49:37.790 --> 00:49:39.600
Regev Nathansohn: Thank you.

237
00:49:40.090 --> 00:49:44.590
good afternoon. It's great to be here. Do you hear me?

238
00:49:44.710 --> 00:49:49.649
Yeah, you can share your screen, your presentation if you

239
00:49:49.750 --> 00:49:52.209
he one.

240
00:49:54.840 --> 00:49:56.820
Regev Nathansohn: Okay, see this.

241
00:49:57.500 --> 00:49:58.270
Yep.

242
00:49:58.870 --> 00:50:02.429
Regev Nathansohn: excellent. Just Hi this morning.

243
00:50:03.500 --> 00:50:04.310
Regev Nathansohn: And

244
00:50:06.480 --> 00:50:17.860
Regev Nathansohn: okay, it's really great to be here in this joint session of my 2 favorite groups. so thank you for

245
00:50:18.050 --> 00:50:25.969
Regev Nathansohn: managing this session. And thank you, Gary, my call organizer, and I'm really sorry for not being there with you this moment.

246
00:50:26.110 --> 00:50:29.539
Regev Nathansohn:  my presentation today

247
00:50:30.100 --> 00:50:31.770
Regev Nathansohn: may contain

248
00:50:32.120 --> 00:50:37.670
Regev Nathansohn: information that some people might find upsetting.

249
00:50:39.210 --> 00:50:49.729
Regev Nathansohn: Wait. how can this be determined? Such announcements are made by digital algorithms today. but they are clearly not. New.

250
00:50:50.270 --> 00:51:04.010
Regev Nathansohn: Human cultures are based on what we may call cultural algorithms, meaning the shared values regarding what is commonly considered legitimate and what might be upsetting.

251
00:51:04.780 --> 00:51:07.840
Regev Nathansohn: So what are digital algorithms

252
00:51:08.720 --> 00:51:27.579
Regev Nathansohn: and simply put digital algorithms, our computational procedures that are programmed to process digital data today, it is mostly big data based on conditional. If, then, sentences that may relate to culture as well as to other fields

253
00:51:28.150 --> 00:51:43.600
Regev Nathansohn: in order to talk today about the role of digital algorithms in visual culture meaning about the growing role in the trialectics of production, circulation and interpretation of visual images.

254
00:51:43.690 --> 00:51:58.440
Regev Nathansohn: we should revisit the pre computational era in visual culture. When analogue algorithms dominated the cultural industry. Going back several decades will allow us to better see

255
00:51:58.450 --> 00:52:02.200
Regev Nathansohn: what has changed and what remained the same

256
00:52:02.750 --> 00:52:18.630
Regev Nathansohn: by genealogizing the algorithmic aura in visual culture. I will briefly argue here that contemporary desires for algorithmic intervention in the visual culture, in the visual culture, predate the digital age

257
00:52:18.700 --> 00:52:29.949
Regev Nathansohn: and go back to the invention of the camera as a technology capable of capturing reality or part of reality, supposedly in an objective way

258
00:52:30.070 --> 00:52:40.529
Regev Nathansohn: in Fox Talbot towards from 1,844. Today's algorithmic aura reminds his perception of photography as the pencil of nature.

259
00:52:42.060 --> 00:52:54.799
Regev Nathansohn: Critical thinkers from Du Bois, Benjamin Virgil tag, and others to contemporary critical, algorithmic scholars taught us that the social practice of codifying data in images

260
00:52:54.800 --> 00:53:12.820
Regev Nathansohn: mit Ctl and is intertwined with socio-cultural classifications, such as race, gender class, and other categorizations that are perceived as having visible attributes while always hiding other sorts of data, thereby turning them invisible one

261
00:53:13.540 --> 00:53:26.749
Regev Nathansohn: today the ability to visually capture reality, to extract and identify data. to take actions based on such data or to create virtual realities, is expanding.

262
00:53:26.840 --> 00:53:35.309
Regev Nathansohn: having higher, although not necessarily more accurate performances, resolutions operating at a greater speed.

263
00:53:35.440 --> 00:53:41.740
Regev Nathansohn: and influencing more spheres of the daily lives of a growing number of people around the globe.

264
00:53:42.960 --> 00:53:54.230
Regev Nathansohn: Simultaneously, the role of in the trialectics of production, circulation and interpretation of images is expanding and developing its own aura

265
00:53:55.130 --> 00:53:58.400
Regev Nathansohn: before further analyzing what happens today

266
00:53:58.780 --> 00:54:10.850
Regev Nathansohn: and how it affects our subjectivity. Let me go back to research. I conducted 20 years ago with pressed photographers right when they started using digital cameras.

267
00:54:11.260 --> 00:54:18.240
Regev Nathansohn: Looking back at my findings and arguments. I can phrase them today in a more contemporary terminology.

268
00:54:18.820 --> 00:54:23.939
Regev Nathansohn: What I saw then is that when press photographers enter a scene.

269
00:54:24.050 --> 00:54:37.020
Regev Nathansohn: they strive to produce the photograph that would later be selected by editors, knowing that what editors select is what they think would be easily and quickly understood by the potential viewer.

270
00:54:37.820 --> 00:54:43.390
Regev Nathansohn: 20 years ago I called it banal mode of image production.

271
00:54:43.420 --> 00:54:50.069
Regev Nathansohn: because it described the production or better reproduction of what is already known.

272
00:54:50.440 --> 00:55:09.980
Regev Nathansohn: There are no surprising images in this mode of production. These images show us what we already have in our shared visible knowledge. We know what the world looks like, we know what the demonstration looks like, and we know or think we know, what environmental crises look like

273
00:55:10.660 --> 00:55:21.060
Regev Nathansohn: analyzing hundreds of photographs taken in the context of the Israeli occupation of Palestine revealed that most images showed nothing new.

274
00:55:21.540 --> 00:55:27.479
Regev Nathansohn: Grave, as it were, and still is. These images show the banality of colonialism.

275
00:55:28.110 --> 00:55:46.909
Regev Nathansohn: When I interviewed press photographers and observed how they do what they do. I learned the logic of their practice in every situation recognized as an event. Press photographers see their role as creating the frame that will quickly and effectively tell the whole story.

276
00:55:47.300 --> 00:56:04.810
Regev Nathansohn: Such photographic practice is usually learned informally, not in an institutionalized way, but through mimicking on the job of the job training, professional isomorphism, and learning from what proved to be a successful entry into the front and home pages.

277
00:56:05.730 --> 00:56:16.050
Regev Nathansohn: Today, we can call this mode of production and analogue algorithmic photography, using non-digital. If, then, conditional practices.

278
00:56:16.900 --> 00:56:28.769
Regev Nathansohn: these best photo of photographers also produced a small number of photographs with high aesthetic value. Without focusing primarily on the news events they were sent to document.

279
00:56:28.920 --> 00:56:39.379
Regev Nathansohn: In my interviews they explained that this session of production was directed mainly to create images for art exhibitions and not for news pages.

280
00:56:39.710 --> 00:56:45.940
Regev Nathansohn: Such images are analyzed by Walter Benjamin Roland, Bart and Susan Black Moors, and others.

281
00:56:46.030 --> 00:56:53.369
Regev Nathansohn: Another form of depoliticizing, by overstimulating the senses of the viewers.

282
00:56:54.030 --> 00:57:07.280
Regev Nathansohn: A similar aesthetic mode of production is what we can see today in images produced by Ais such as Dali or Dali, 2, Mid Journey, and even today by Photoshop Beta.

283
00:57:08.780 --> 00:57:22.800
Regev Nathansohn: my main research question, 20 years ago, focused on the social practices of producing critical images of military occupation ended up finding such images in a surprising place

284
00:57:23.130 --> 00:57:33.909
Regev Nathansohn: in private photo albums of Israeli-combat soldiers who patrolled the castle of Hebron, holding a weapon in one hand and a camera in another.

285
00:57:34.510 --> 00:57:46.710
Regev Nathansohn: Some of the images they produced showed unclear social situations. What these images show does not correspond with the common visual knowledge about the military occupation.

286
00:57:46.950 --> 00:57:55.540
Regev Nathansohn: Such images push the viewers to ask questions about what they see. Instead of confirming what they already know.

287
00:57:55.910 --> 00:58:03.589
Regev Nathansohn: Following the laz and wateri. I saw these images as depicting the rezomatic reality of colonialism

288
00:58:03.650 --> 00:58:11.349
Regev Nathansohn: resulting from resulting from an unintentional, transgressive mode of visual production.

289
00:58:11.990 --> 00:58:17.650
Regev Nathansohn: Today. We can call this mode of production none. Algorithmic photography.

290
00:58:19.030 --> 00:58:28.940
Regev Nathansohn: The 3 modes of visual production operate in the trileactics of visual culture under the growing power of large-scale corporations.

291
00:58:29.910 --> 00:58:39.529
Regev Nathansohn: all 3 modes of production, the transgressive, the aesthetic, and the banal show different relations between production, circulation and interpretation.

292
00:58:40.370 --> 00:58:55.669
Regev Nathansohn: If we focus primarily on the banal mode of production of an a local borithmic photography. It shows a tight toilet whereby the expected interpretation of images feeds the production. Assuming a certain circulation

293
00:58:55.990 --> 00:58:57.989
Regev Nathansohn: in the early 2,000.

294
00:58:58.160 --> 00:59:09.499
Regev Nathansohn: The translate, the transition to digital photography gave more power to the photographers, both with regards to their control over the production of images.

295
00:59:09.970 --> 00:59:16.149
Regev Nathansohn: They did not have to count on dark rooms, and could improve on the ground images that were not good enough.

296
00:59:16.270 --> 00:59:20.999
Regev Nathansohn: and they could control the selection of images they sent to their editors.

297
00:59:21.240 --> 00:59:38.489
Regev Nathansohn: And some 20 years later, when we ask whether there is an algorithmic turn in visual culture, I think that the question is less about the algorithms and more about how changes in human subjectivity affect the toiletics of visual culture

298
00:59:40.130 --> 00:59:51.299
Regev Nathansohn: rephrasing the arguments of my research. I try to show how algorithmic thinking these, if, then, conditional practices was always there.

299
00:59:51.330 --> 01:00:07.030
Regev Nathansohn: I want to suggest here that what may be identified as the digital algorithmic turn in contemporary visual culture receives it over its aura by the d subjectification of visually based social practices.

300
01:00:08.530 --> 01:00:20.779
Regev Nathansohn: The perceived neutrality of algorithms shifts the responsibility of its consequences from the human subject to the supposedly non-human automated and learning machine.

301
01:00:22.090 --> 01:00:31.470
Regev Nathansohn: According to cite a fell, a computational photography moves away from images composed from a single temporal frame.

302
01:00:31.490 --> 01:00:41.669
Regev Nathansohn: In Google Pixel 3, for example, the process of taking a photograph is transformed from a human operator capturing a moment of choice

303
01:00:41.910 --> 01:00:55.539
Regev Nathansohn: to a process of computational capturing and processing of a stream of 15 frames to create an image with greater resolution and lower levels of noise than would be possible with a single exposure.

304
01:00:57.520 --> 01:01:09.010
Regev Nathansohn: It's one photographer told Doron. Alter rats in a recent K that published article in photographies. His self-perception is minimized to that of a human tripod.

305
01:01:09.250 --> 01:01:15.510
Regev Nathansohn: According to him, the camera turned his body into an agent of image making software.

306
01:01:16.880 --> 01:01:24.120
Regev Nathansohn: Another expression expression of digital algorithms recently analyzed in an article by Yushin Zhang

307
01:01:24.160 --> 01:01:27.350
Regev Nathansohn: is, have you weighs moon mode

308
01:01:27.420 --> 01:01:40.789
Regev Nathansohn: part of what the company calls its master? AI. A pre installed algorithmic program that intelligently detects and enhances images of the moon captured by the user's phone.

309
01:01:41.800 --> 01:01:43.560
Regev Nathansohn: According to Zhang.

310
01:01:44.050 --> 01:01:51.730
Regev Nathansohn: always algorithmic in intermediation, recontextualizes, photographic practices.

311
01:01:51.860 --> 01:02:02.659
Regev Nathansohn: She argues that algorithmic photography is a kind of dynamic multi-scaler process of mediation through which human and mechanic agents

312
01:02:02.680 --> 01:02:08.130
Regev Nathansohn: are entangled to negotiate and claim the visuality of the moon.

313
01:02:09.590 --> 01:02:24.129
Regev Nathansohn: These examples show how the subjectivity of the human image maker is minimized, and how the expansion of digital algorithms, Titans, the tri elective of the production, circulation and interpretation.

314
01:02:25.310 --> 01:02:30.560
Regev Nathansohn: It therefore becomes crucial to trace the genealogy of such algorithmic practices.

315
01:02:30.610 --> 01:02:47.210
Regev Nathansohn: to re-subjectify the human agency in this algorithm return to open the new possibilities for accountability transparency responsibility and equitable algorithmic interventions in the cultural production of images

316
01:02:47.450 --> 01:03:01.590
Regev Nathansohn: simultaneously. It is crucial to also evaluate in this context the growing return to various analog visual practices and to other forms of image making that allow for critical ways of seeing

317
01:03:01.650 --> 01:03:07.199
that can flourish in the gaps between production, circulation and interpretation.

318
01:03:07.720 --> 01:03:15.169
Regev Nathansohn: the more algorithms are being programmed to interact with each other. The closer these components become.

319
01:03:15.350 --> 01:03:25.249
Regev Nathansohn: the closer these components become, they can better serve profit and conservative purposes, leaving less space for sensitive content.

320
01:03:25.260 --> 01:03:44.079
Regev Nathansohn: meaning, less space for human intention and interaction, less based for difference, for skill, for creativity, and for trial and error, in other words, less space for subjectivity and for alternative and radical ways of showing and seeing.

321
01:03:44.430 --> 01:03:45.330
Regev Nathansohn: Thank you.

322
01:03:53.070 --> 01:04:06.419
Thank you very much. Thank you. Rakov. The last speaker in our session is Matthias Zomer from the Ham meets University of Technology in Germany

323
01:04:06.500 --> 01:04:12.050
and smart photos in algorithmic situations.

324
01:04:13.890 --> 01:04:28.989
Matthias Sommer: Yes, thank you very much for the invitation and organization of the panel. I hope you can hear me. My contribution can connect to much of what has already been said, but I also hope to add a few new aspects to the discussions.

325
01:04:29.050 --> 01:04:34.940
Matthias Sommer:  yeah, the title of my presentation is smart photos and algorithmic situation.

326
01:04:34.950 --> 01:05:04.709
Matthias Sommer: exploring the relationship between personal image compilations and algorithmic image clusters. As you can gather from the title of my talk with focus on 3 concepts, the smart photo problem, personal image compilations and algorithmic image clusters. So what is as a smart photo problem in 2,016 that introduced a function called smart photo as a recap tinder is an application to find potential dating partners

327
01:05:04.910 --> 01:05:06.350
Matthias Sommer: to match up

328
01:05:06.670 --> 01:05:07.380
this

329
01:05:07.570 --> 01:05:28.219
Matthias Sommer: to match individuals. Every user creates a profile with up to 9 image that provide information about their personality. Initially, users arrange their image in a fixed ranking from one to 9 when shown to another user, the first image of a given candidate is depicted, and users must decide whether to skip to the next one, swipe left or right.

330
01:05:28.220 --> 01:05:39.200
Matthias Sommer: or view more pictures by the same user tap up or down. Consequently users have to determine which image they want to be displayed. First, in comparison to other user's content.

331
01:05:39.200 --> 01:05:54.809
Matthias Sommer: The smart photo algorithm offers option to dissolve the fixed structure of the image composition, instead of the individual deciding which image image presents the best, almost click-worthy version of themselves, they delegate the ranking of the order to an Ivorism

332
01:05:55.160 --> 01:06:21.480
Matthias Sommer: when addressing tinder. My intention is not to delve into the sociology of online dating. Instead, I argue that the smart photo problem can be perceived as a central issue wherein algorithmic processes are not merely abstract or technological concepts, but integral components of individuals. Common sense constructions in their daily lives. My interest lies in the everyday manifestation of this algorithmic turn, particular in relation to image use.

333
01:06:21.820 --> 01:06:32.629
Matthias Sommer: In 2,020, a viral trend emerged on social media known as the Dolly Pardon challenge. This trend was initiated by the musician, who posted 4 different images of herself.

334
01:06:32.630 --> 01:06:53.009
Matthias Sommer: The concept behind the challenge was simple. Present for versions of one self that answers the question, Who am I? Instead of linking this question to specific situations or social roles, for example, who am I as a teacher, mother, or football player. Participants were tasked with using 4 different profile photos corresponding to different social platforms.

335
01:06:53.010 --> 01:07:16.419
Matthias Sommer: link it in Facebook, Instagram and Tinder. The mean helps to visualize the, for instance, of identity work in digital environments. The only part is comment. Get you a woman who can do it all further emphasize the challenge associated with creating a carefully constructed persona. I put Baker and Rojack here that is not only aspirational but also relatable, emulatable, and accessible

336
01:07:16.490 --> 01:07:43.229
Matthias Sommer: the problem of integrating the multitude of social circuits that individuals are part of is not a new phenomenon. Similarly, a role conflict such as the tension between a good father and a good employee are not novel experiences, however, in contrast to face-to-face relationships, the way situations are defined, changes in algorithmic environments. Moreover, an increasing amount of identity. Construction takes place using visual material.

337
01:07:43.430 --> 01:08:07.310
Matthias Sommer: and some Strauss describes how, in a problematic situation, it is not sufficient to identify the current other. But one must Perry person identify his current cell in regard to image use and algorithmic situation, we have to make assumptions or guesses about the grounds of others. Actions in the form of image use in these situations, whereas modes of interaction can change in algorithmic situations.

338
01:08:07.310 --> 01:08:17.159
Matthias Sommer: practices of image use transform as well with the rise of Instagram, drawing attention to oneself through image compilations, has become an ambitious phenomenon.

339
01:08:17.160 --> 01:08:37.690
Matthias Sommer: Unlike application photos, funeral portraits, or traditional family photography, the focus is no longer on selecting a single photograph to depict the character of the personality or capture the essence of a family unit. Instead, the aim is to create personalized clusters, a digital persona consisting of a series of pictures

340
01:08:37.690 --> 01:08:51.189
Matthias Sommer: in contrast to pictures, albums on Facebook, which are created for specific topics or occasions, and make accessible to a selected group of users. The creation of a photo block is not the primary focus on the mentioned social platforms.

341
01:08:51.320 --> 01:09:03.409
Matthias Sommer: although a personal image cluster is created and new images are added to it. These images are primarily targeted towards algorithmic image clusters created or facilitated by the platforms.

342
01:09:03.410 --> 01:09:21.770
Matthias Sommer: The smart photo problems, Homer argument we encounter in situations in which we present ourselves with the help of image in algorithmic image clusters. It is a temporal visual answer to the question, what is going on? An answer to identify one's current self in a given situation.

343
01:09:22.270 --> 01:09:43.449
Matthias Sommer: I use the image cluster analysis developed by Michelle Miller for analyzing the compositional and montage principles of image clusters. Miller emphasizes the importance of image comparison for the understanding of visual phenomena. He defines image clusters as a total use of photography in which images are compiled to form a greater whole.

344
01:09:43.450 --> 01:10:01.710
Matthias Sommer: In the following discussion I will examine the 3 ideal types of image clusters developed by Miller and apply them to personal image compositions and algorithmic image clusters the central distinction between these types license, a comparison criteria used to compile and arrange the image. Corpus

345
01:10:01.760 --> 01:10:09.849
Matthias Sommer: narrative image clusters are characterized by criteria of sequential coherence with respect to temporality and finality.

346
01:10:10.080 --> 01:10:19.269
Matthias Sommer: Justificatory clusters, on the other hand, involve image compiled and compiled, based on a logic of subsumption, conceptual or numeric criteria.

347
01:10:19.380 --> 01:10:25.040
Matthias Sommer: These clusters organize image according to specific categories or conceptual frameworks.

348
01:10:25.310 --> 01:10:37.890
Matthias Sommer: In iconic image clusters. The compiled image are selected, based on their criteria of visual similarity and differences, thereby creating iconic visual values and representational forms.

349
01:10:38.380 --> 01:10:42.730
Matthias Sommer: Let me start with one example of narrative use of image composition.

350
01:10:42.760 --> 01:10:59.520
Matthias Sommer: What you can see here is an image compilation taken from the one-year performance by Deutsche, which he started in April 1,980. The artist explains the performance like this. I shall punch a time clock in my studio every hour, and on the hour for one year

351
01:10:59.620 --> 01:11:06.869
Matthias Sommer: with a 16 moving camera. I shall document each time I punch the time clock by shooting one

352
01:11:07.020 --> 01:11:08.020
Matthias Sommer: rain.

353
01:11:09.150 --> 01:11:28.839
Matthias Sommer: The image compilation serves as documentation of the one-year performance. Although the setting and a static of the image remain consistent, they depict the progression of time. This is a shift through various elements, such as capturing the growth of the here over time and incorporating stem time cards for each day.

354
01:11:30.340 --> 01:11:38.480
Matthias Sommer: The project list, with more from the artist Christian Boltonski, can be used as an example for the classificatory use of image compilations.

355
01:11:38.530 --> 01:11:43.290
Matthias Sommer: The slide shows part of the collection of images that he used for his project.

356
01:11:43.400 --> 01:11:53.060
Matthias Sommer: all compiled image. In the catalogue there are more than 600 black and white, and illustrations are cut out. Portraits from orbitories.

357
01:11:53.290 --> 01:11:58.920
Matthias Sommer: The title of the catalogue of the dead. Swiss gives the image combination its specific meaning.

358
01:12:00.150 --> 01:12:15.529
Matthias Sommer: Finally, I will present 2 examples of iconic image clusters, futuring pictures of individuals, one of which directly relates to the question of algorithmic image justice. The first example is, the photo block holds still a portrait of our nation in 2,020.

359
01:12:15.540 --> 01:12:28.370
Matthias Sommer: The project was the result of a photo competition initiated by the National Portrait Gallery in London, Uk. The gallery called upon the public to submit photographs capturing the essence of everyday life. During the pandemic

360
01:12:28.580 --> 01:12:35.820
Matthias Sommer: over 31,000 submissions were received from these submissions. A collection of 100 image was chosen by a jury

361
01:12:35.970 --> 01:13:01.740
Matthias Sommer: through the exhibition. Individual portraits are just a post creating a powerful composite image of a collective entity, a nation confronting a crisis so selected image within the photo block, with their shared and contrasting features, contribute to the creation of a compelling Heapa image that conveys a new meaning rather than depicting personal distress the image, compilation portraits, the collective struggle, experience.

362
01:13:01.760 --> 01:13:07.380
Matthias Sommer: It highlights a community united in isolation, bound together by shared.

363
01:13:07.780 --> 01:13:35.070
Matthias Sommer: The second example is a screenshot from an Instagram profile, which serves as an illustrative case of how images are created and arranged, based on similarities and differences, creating a compelling narrative for a visual persona. In this case a dedicated runner. The profile demonstrates a specific image, aesthetic, that is consistently applied throughout the image cluster in the top right corner of each image, the logo of a company producing variable technology for runners is displayed.

364
01:13:35.130 --> 01:13:41.739
Matthias Sommer: The individual often presents herself to full body portraits captured from behind with the road in view.

365
01:13:41.790 --> 01:13:52.990
Matthias Sommer: Additionally, the image frequently emphasize her running shoes, feed, and leg muscles. The feet are depicted in a manner that conveys movement with one foot appearing to lift off from the ground

366
01:13:53.140 --> 01:14:19.139
Matthias Sommer: the difference between the 2 outlined iconic image clusters are instructive, considering smart photos and algorithmic image clusters. In the first example, portraits of individuals were submitted to a prestigious art Museum with aspiration of having one's portrait selected for exhibition. Although no algorithmic situation was involved, the issue of visibility and inclusion and exclusion was mediated through a selection process conducted by a jury.

367
01:14:19.150 --> 01:14:31.940
Matthias Sommer: In the case of the Instagram profile, we enter the domain of algorithmic situations. Each image not only becomes part of the personal image compilation, but also targets. Multiple algorithmic image clusters.

368
01:14:32.300 --> 01:14:56.709
Matthias Sommer: This slide shows one picture of the personal image compilation displayed in the explorer feet. The explorer feet is a combination of posts tailored to the users. Interest and behavior. It showcases potentially interesting content. In this case the explorer fee demonstrates how the image is positioned among symmetrically similar image, while also being combined with images from other subject areas.

369
01:14:57.090 --> 01:15:18.500
Matthias Sommer: The hashtag feat is another important algorithmic cluster. When users search for a specific hashtag on Instagram, they have shown a feed of all posts that use that hashtag. This feature enables users to find content that aligns with their interests. Additionally, users can choose to follow hashtags which affect the content displayed in the discovery and home feeds

370
01:15:18.600 --> 01:15:25.100
Matthias Sommer: from the perspective of someone posting an image. Hashtags are used to engage with these algorithmic image clusters.

371
01:15:25.140 --> 01:15:35.309
Matthias Sommer: The original post used 30 hashtags. either. You can see. no, yes, here you can see 3 of them displaying the same image from the Instagram profile.

372
01:15:35.520 --> 01:15:57.400
Matthias Sommer: and because of the time I'd skip the part where I go into more detail about the differences between the cluster while using the same technique and quite definition functioning as classificatory clusters. What you notice at first lens, is that it's semantic. Hashtag, like running on the left, puts the image in a completely different context than the hashtag storms in the middle, or nurse life on the right

373
01:15:57.730 --> 01:16:17.609
Matthias Sommer: to better understand the relationship between personal image combinations and algorithmic image cluster. I use the concept of mediatized self advertisement. The auxiliary analyzed marriage advertisements as a medium created for personal selection according to similar self advertisement, emerged as a consequence of the individualization of modern personality.

374
01:16:17.650 --> 01:16:23.790
Matthias Sommer: This form of advertising serves the purpose of making specific aspects of oneself. Visible

375
01:16:23.970 --> 01:16:46.719
Matthias Sommer: marriage advertisements as a specific selection tool facilitate distinct forms of competition among individuals. The so-called personal selection and newspapers historically provided infrastructures like social platforms today for various personal announcement, including looking for romantic partners, friendship, companionship as well as announcement of birth or death of persons

376
01:16:46.720 --> 01:16:54.359
Matthias Sommer: in regard to the question what qualities of a person could be stated in an advertisement similar identified one-centered problem.

377
01:16:54.360 --> 01:17:14.540
Matthias Sommer: He stated that only the financial situation of an individual could be precisely conveyed through such ads. He notes, I quote all other qualities of the person, such as their outward appearance, character, affability, and intelligence cannot be accurately described in a manner that creates a clear and direct individual interest.

378
01:17:14.590 --> 01:17:27.139
Matthias Sommer: Hence, as similarly regarded, the financial status of an individual as a primary criterion in marriage advertisements, and as an implicit comparative criterion, the advertisements were making persons comparable with

379
01:17:27.520 --> 01:17:53.269
Matthias Sommer: modern selection tools have significantly transformed this dynamic if a illus argues that today there is a specific form of Internet imagination, a form of imagination that is based not on the body, but on linguistic, exchanged and textual information for illos Internet imagination is information sick? It is, I quote, dominated by a verbal overshadowing the prevalence of language in the process of evaluation.

380
01:17:53.430 --> 01:18:20.200
Matthias Sommer: The creation of digital personas with algorithmic image clusters shows that an Internet imagination is more and more based on visual practices using personal image compilations. It could be argued that in visual personal selection tools, one capabilities to present an outward appearance give visual representations of one's character and the ability to present oneself in emotional categories from the foundation of self presentation.

381
01:18:20.340 --> 01:18:31.770
Matthias Sommer: Personal image clusters on tinder are in so far information sick that every new image adds information to the creative persona. Following Evan Goffman, a visual personal selection tools

382
01:18:31.770 --> 01:18:52.740
Matthias Sommer: cannot be understand in their potential to create text-based knowledge alone. In gender advertisement, government emphasizes the visual competency of the I, and notes that through image compilations one can introduce aspects into consideration, even if on his in surface, in sufficient literary skill to summon them up through birds alone.

383
01:18:53.320 --> 01:19:10.380
Matthias Sommer: Einstein.

384
01:19:10.390 --> 01:19:30.780
Matthias Sommer: But more than that, a user gets in contact with other iconic image clusters here. Potential partners in form of the algorithmic image cluster further, more users are anticipating in an imaginary, in an imaginative way, or through looking at them directly, image of competing users with which one's own image are just opposed with

385
01:19:30.880 --> 01:19:39.819
Matthias Sommer: in regard to image use. The engagement in algorithmic clusters can be described as a social practice of imagination to borrow a term by Ion upper durai

386
01:19:39.910 --> 01:20:04.730
Matthias Sommer: about durian notes. How these practices are characterized by different conceptions of the image. In the case of tinder, we have digital image in the form of the personal image compilations uploaded by the user. It's the same time algorithmic clusters can create new forms of imagined communities in which social verification of authenticity oriented authenticity occurs.

387
01:20:04.810 --> 01:20:22.650
Matthias Sommer: Finally, this social sociality in algorithmic image clusters can create specific forms of imaginaries. For example, concept of good personalities in regard to specific social relationships. In the case of tinder, the focus on self advertisement established specific ways of seeing one cell and others

388
01:20:22.650 --> 01:20:47.899
Matthias Sommer: writing about about publicity image in city centers. John Berger emphasizes our publicity image while competing with each other, create a language in itself. They, I quote, confirm, and enhance each other. They propose to each of us that we can transform ourselves or our lives. This statement also, what's true for self promoting image within algorithmic image justice on dating acts.

389
01:20:47.900 --> 01:20:54.469
and allow me to conclude my presentation with 4 Caesars regarding smart photos within algorithmic image clusters.

390
01:20:54.470 --> 01:21:19.010
Matthias Sommer: First, algorithmic situations in visual selection tools are established through the combination of personal and algorithmic image clusters. Second, in these situations different modes of interaction are combined, especially forms of mediated crazy interaction in the form of personal image compilations and mediated online interaction in the form of algorithmic image clusters.

391
01:21:19.480 --> 01:21:44.730
Matthias Sommer: To better understand the relationship between algorithms and visual social communication. We need to focus on the social construction of algorithmic image use. And finally, the smart photo problem highlights the significance of algorithms as social practices of imagination. The focus is not solely on algorithmic clusters as a specific entity, but also on how they are embedded in the everyday use of image.

392
01:21:45.170 --> 01:21:47.730
Matthias Sommer: Thank you very much for your attention.

393
01:21:53.420 --> 01:21:59.449
Thank you very much, Matthias, and thank you to all of

394
01:21:59.550 --> 01:22:00.880
speakers.

395
01:22:00.940 --> 01:22:05.430
yeah, yeah, sure.

396
01:22:05.480 --> 01:22:08.070
Hello.

397
01:22:08.520 --> 01:22:22.710
so let's let's turn to. next next phase, and I'm really open to hear your questions to speakers.

398
01:22:24.730 --> 01:22:25.950
I've got

399
01:22:28.860 --> 01:22:32.959
Hello, I want to do a question for

400
01:22:33.180 --> 01:22:39.459
the first one and the last one. Yeah.

401
01:22:40.020 --> 01:22:45.729
throughout the development of social media. We had the creation of what we call the newsfeed

402
01:22:45.800 --> 01:23:09.309
in 2,006. Before that I feel social media. The experience of social media was completely different, because through the new news speed we have the opportunity to have this 5 s of fame that everyone wanted to achieve in the newspapers, but it's now available on a massive scale. And then the algorithms come into play so that

403
01:23:09.340 --> 01:23:13.330
each and everyone can achieve this

404
01:23:13.940 --> 01:23:24.630
social validation of yourself. I wanted to ask if, how would you see the the grow of the news feed into these dynamics

405
01:23:24.770 --> 01:23:28.650
of algorithm relationships?

406
01:23:32.080 --> 01:23:33.719
Do you want me to start? Thank you.

407
01:23:34.090 --> 01:23:35.090
whatever.

408
01:23:35.390 --> 01:23:52.859
 well, I can only just reflect on some of the things that are participants in the study talked about, and people were saying said things like, Oh, the algorithm doesn't love me anymore. Because

409
01:23:53.340 --> 01:23:58.479
after a while, what they, what they notice are the changes in the behavior of the news feed.

410
01:23:58.660 --> 01:24:08.409
And they talk to each other, and they see that things that they expect to show up don't show up. They don't see their friends anymore. And I mean. You know.

411
01:24:09.210 --> 01:24:16.779
it's interesting because it's so complex. There are people that start to feed in, to try to hack into it, to manipulate it to.

412
01:24:17.160 --> 01:24:38.570
you know, hashtag it, and I was particularly this in the last presentation, the 30 hashtags for the you know that that was really sort of embarrassing for millennials. I mean, they don't want to look like they're trying to get attention on the news feed. They don't want to be featured on the explorer page, because it looks like they're trying too hard.

413
01:24:38.600 --> 01:24:45.030
So they've mastered things like putting hashtags in and then deleting them

414
01:24:45.400 --> 01:24:56.170
and they do all kinds of tricks to try to manipulate or deal with the news feeds. So I think that there's an ongoing tension

415
01:24:56.270 --> 01:25:12.389
between how people deal with the newsfeed, and I don't know if that answers your question, but there's definitely a people are like, you know, like the news feed. But More and more. you know, amongst millennials from what we talked about in 2 Canadian cities.

416
01:25:12.880 --> 01:25:37.039
 addressing people, university age from 4 different universities is that they're disengaging in a lot of ways like they're they're not looking. They don't trust images the way different generations trust images. They're always suspicious of the image. They're always like looking for watermarks or ways that people change the image to see if you know what happened here. Stuff like that.

417
01:25:40.500 --> 01:25:46.599
Matthias Sommer: yes, maybe maybe I I hope you can give me. Maybe I can add to this.

418
01:25:46.620 --> 01:25:54.139
Matthias Sommer: I think it's a it's a difficult question to answer in in this general general way, how the newsfeed changes image use. But

419
01:25:54.180 --> 01:26:19.179
Matthias Sommer: in my part, I really think that the combination, what I, what I said between this personal image clusters and this algorithmic clusters changed by more focusing on this algorithmic way of pottering oneself, as I said, compared to, for example, in the old in the old times with Facebook and albums I created for peer group

420
01:26:19.180 --> 01:26:44.179
Matthias Sommer: or something. I think maybe 1. 1. One more aspect which I found interesting is the whole discussion about algorithms as a black box. Right? This. If you if you talk about Tik Tok or Instagram that, I believe isn't change. And he doesn't like me, as we heard before. This, I think, is really interesting, because nowadays we have this ad hoc see rising that the algorithm in every

421
01:26:44.180 --> 01:27:12.659
Matthias Sommer: life. As I said for me, it is very interesting to focus on this algorithmic term that people are trying to anticipate what the algorithm is doing. And actually, it's not so important that I know the whole structure of the algorithm, like the like, the Thomas phenomenon. And if if I I think the algorithm discriminates me. It's it's almost the same right? And and that, I think, is is quite important. This kind of you rising algorithmic structures because of the news feed.

422
01:27:12.800 --> 01:27:14.230
Matthias Sommer: yeah.

423
01:27:15.880 --> 01:27:39.070
Matthias Sommer: But but then I really think we need to study more carefully, especially if we talk about the hashtags, how they are really applied in given image practices and communities. Because I really think that they are a lot of differences. If we study, for example, hashtags more used symmetrically to to get visibility or hashtags used in in subcultures, etc. So

424
01:27:39.070 --> 01:27:48.259
Matthias Sommer: I really think that it doesn't have to be this one news algorithm on Instagram, I think even there we have to differentiate.

425
01:27:51.580 --> 01:28:02.389
Can I just add one more thing? I don't know if this addresses what you were thinking about. People are curating a lot more who? Which are the groups that they include

426
01:28:02.650 --> 01:28:13.469
so that they can. They curate who belongs in their social media groups. and it helps manage the feed.

427
01:28:13.630 --> 01:28:19.530
and they also curate their own. What it looks like overall as you saw, so I don't know if that

428
01:28:19.830 --> 01:28:21.029
it it up.

429
01:28:22.490 --> 01:28:32.679
Thank you. To all of you a very interesting presentation, looking at different aspects of the algorithmic age in which we live. But

430
01:28:32.830 --> 01:28:39.720
I want to ask you, Caroline, a question with respect, a more general theoretical

431
01:28:39.930 --> 01:29:07.120
issue going back from your presentation to your book. Do you really believe that we are at the end of photography because we already went through bars, song, tang, and so on. Personally, I really do believe that we are experiencing photography is never we did in our short human history, connect to development

432
01:29:07.120 --> 01:29:33.229
of photography from the camera. Oscar, can you please tell me something more? Thank you.

433
01:29:33.230 --> 01:29:54.700
In photography like photography, is no longer a memory. It's sort of a an ephemera, right? People don't look back on their selfies from last year. I mean, they really want those to disappear. It's the only one that counts as the now, right? So that's an epistemological shift. So photography has had that shift, that's all it is. Thank you.

434
01:29:57.250 --> 01:30:01.770
 Hi, question for reggae.

435
01:30:01.810 --> 01:30:04.800
 So

436
01:30:05.270 --> 01:30:17.430
this in salo in Sauberong, removing this era of magic from from algorithm and stressing the continuities is absolutely important in it

437
01:30:17.570 --> 01:30:33.770
reminds me of the work of Massimo Iiraldi on on the, on the digital habitats. When it suggests that algorithms basically internal internalize the social world in a similar way that we do.

438
01:30:34.130 --> 01:30:47.109
Yet I want to challenge you to think more, to elaborate a bit more about the differences, for example, about the way that Gen. AI

439
01:30:47.510 --> 01:31:11.879
reflects the field in different time space, and collapse them into a single point, where whereas the a field, such as the journalistic photography field develops over time, and every time we we experience it from a different perspective, or the way the different field position collapses

440
01:31:11.880 --> 01:31:29.879
into a single ideal type in the way these algorithmic models produce images.

441
01:31:30.030 --> 01:31:31.370
The Islamic.

442
01:31:33.440 --> 01:31:36.630
Regev Nathansohn: Thank you. Already. Right? I recognize the voice.

443
01:31:39.320 --> 01:31:40.640
Regev Nathansohn: And

444
01:31:41.090 --> 01:31:43.979
Regev Nathansohn: it's a great question. I

445
01:31:44.320 --> 01:31:46.989
Regev Nathansohn: I agree that there are differences.

446
01:31:47.100 --> 01:32:02.080
Regev Nathansohn: and it was really interesting. Only this morning I got a notification on a new article published, and I remember exactly which June I was talking about the social unconscious in AI,

447
01:32:02.300 --> 01:32:11.929
Regev Nathansohn: which which my with and that might explain a little bit, the the the aura that we have around algorithms that it's somehow

448
01:32:11.980 --> 01:32:24.830
Regev Nathansohn: and hides or develops our our collective, unconscious that we cannot reach. manually, analogically.

449
01:32:25.160 --> 01:32:41.290
Regev Nathansohn: And this is something that we may conceive as different. and it also relates to what you talked about. It's a temporality. It's not the the here and now. It's something that develops over time and keeps

450
01:32:41.350 --> 01:32:47.700
Regev Nathansohn: as opposed to what Carolina just said, the the ephemeral, the algorithms know

451
01:32:48.000 --> 01:33:00.050
Regev Nathansohn: they collect, they can refer to past images, they can refer to past classifications, and they can do something more than what each and every one of us can do.

452
01:33:00.110 --> 01:33:10.509
Regev Nathansohn: And I think this is something that we attribute to algorithms as more powerful than what each individual is capable of doing with their own visual devices.

453
01:33:10.700 --> 01:33:11.950
Regev Nathansohn: And

454
01:33:13.940 --> 01:33:18.770
Regev Nathansohn: I think that that that's the answer that I can give you now. But I need to think more about it. Thank you.

455
01:33:23.770 --> 01:33:27.430
Any questions. Yeah.

456
01:33:27.450 --> 01:33:46.290
it doesn't want to anymore. Okay, so I've got a question for Carolina. Basically, I would like to broaden the question asked by a friend here.

457
01:33:46.290 --> 01:34:04.359
Yeah. And have you detected or studied any counter hegemonic practices when it comes to taking selfies.

458
01:34:04.590 --> 01:34:13.409
you know, some subversive forms of counter-self or anything like that.

459
01:34:18.510 --> 01:34:19.380
Yeah.

460
01:34:20.820 --> 01:34:23.300
I didn't directly

461
01:34:24.370 --> 01:34:31.639
is set out to do that. But you know, people do discuss

462
01:34:31.760 --> 01:34:37.329
taking photos of the floor and hashtag it with, you know, ironic hashtags and

463
01:34:37.440 --> 01:34:58.730
So some because people know that people search with hashtags. There is a lot of irony used in the use of hashtags. So it people who use big data, for example, and and scrape twitter. Miss, all that right? So the big data misses all the joking around that's going

464
01:34:58.770 --> 01:35:02.410
 against all this. So I think that

465
01:35:02.810 --> 01:35:06.650
what I saw in terms of subversive or

466
01:35:06.760 --> 01:35:15.029
counter-image practices were turning things into an ironic mode and joking around about it.

467
01:35:15.290 --> 01:35:22.220
But it was interesting that female people who identified as female participants

468
01:35:22.540 --> 01:35:40.039
were a lot more serious and engaged about selfie practice, and people who were identified as male participants were very much more joking about it, and it was not regarded as a serious practice. So in those terms there was a bit of

469
01:35:40.330 --> 01:35:56.929
play with that. But there are other people that are studying sort of small subgroups that create their own visual lexicon around Selfie taking But it's not something that feeds into the logic of Instagram. So that's a really interesting question.

470
01:36:01.410 --> 01:36:08.139
Actually, I also have a question for you, namely, you discussed gender. And I wonder about

471
01:36:08.220 --> 01:36:17.190
the other axis of intersectionality and the way they shape both the algorithmic, imaginary.

472
01:36:17.460 --> 01:36:21.630
and the way people, the young people engage with it

473
01:36:21.700 --> 01:36:43.260
and their norms of representations in relations to which they act. I mean, I know data from from Israel that show very strong class differences in norms of of Instagram representation. I wonder what what we are defining from your study?

474
01:36:46.530 --> 01:36:55.729
Thank you very much for that question. I think we can all agree that the social media platforms are generally really heteronormative.

475
01:36:55.990 --> 01:36:59.770
and so gender non-conforming people

476
01:36:59.830 --> 01:37:16.130
generally find other ways to express or resist or directly opt out. And in our focus groups we had very few people that formed focus groups that identified as gender, nonconforming, and their discourses were completely different

477
01:37:16.360 --> 01:37:20.499
from people that fitting into sort of heteronormative

478
01:37:20.600 --> 01:37:23.969
identification. And we didn't have enough

479
01:37:24.020 --> 01:37:35.970
of those participants to conclusively identify a discourse, and so what we do in the work is just gestured towards it. Yes, there is something.

480
01:37:35.990 --> 01:37:43.039
but we don't know what it is. It hasn't been explicitly addressed, but we saw that it was there, and it was different.

481
01:37:49.870 --> 01:37:51.330
Any questions?

482
01:37:57.680 --> 01:38:00.810
No. don't be shy.

483
01:38:02.570 --> 01:38:04.600
Okay. And

484
01:38:05.450 --> 01:38:07.500
okay, so

485
01:38:08.880 --> 01:38:09.950
we're done.