Symposium proposal

[MarcoViola88]

Hi FictionEro folks, I hope you are doing well!
I apologize for having postponed sketching the intro, I've been drowning in more pressing commitments -- but do not worry, I'm keeping an eye on the literature and I'll be ready to do my task ;-)

But let me put a bold & fun proposal on the table: let's meet physically in Rome, 18-20 September, i.e. let's propose a symposium to the annual conference of the Italian Association of Cognitive Science.
The Call for Paper has been published just a bunch of minutes ago, and I know the organisers well enough so I'm pretty sure they'll love the idea of a symposium on our topics: http://aisc2024.istc.cnr.it/?page_id=21

Other/rather than presenting our current common project, I'm thinking about a symposium called e.g. "Social Cognition and Digital agents", where we could discuss our previous jobs related to e.g. uncanny valley and moral psychology (Antonio), fictional reappraisal (@marcosperduti & @DominiqueMakowski), real is the new sexy.

Quoting from the website:

Rome is always beautiful, and I can ensure that AISC conference are pretty lively, both intellectualy (there's a nice mixture of psychologists, philosopehrs, neuroscientists, computer scientists...) & socially (plenty of informal and outgoing people).
Who's up for it? :-D

[marcosperduti]

Who's up for it? :-D

great idea, I'm in! It would be also a nice opportunity to discuss of future developments.

[Antonio8424]

This is just too good, I´m definetely in!
@MarcoViola88 , would it work for you if I send you an abstract of my proposal so you can integrate it into the description of the proposed symposium? Please let me know what works best for you.

[MarcoViola88]

Happy to see that the proposal picks your interest :-D
Yes Antonio, please do send an abstract draft (200-220 words) at your earliest convenience, I'd then try to weave them all together in a semi-coherent framework :-)
Marco (S.), would you mind sketching an abstract too?

[marcosperduti]

Marco (S.), would you mind sketching an abstract too?

No problem, for when you'll need the abstract?

@DominiqueMakowski in case we present data from the fiction project, do you want to present?

[DominiqueMakowski]

Nah go ahead I'll just ask complicated questions at the end

[marcosperduti]

Nah go ahead I'll just ask complicated questions at the end

Ok, so you'll come just to enjoy Rome :)

[DominiqueMakowski]

[Antonio8424]

Hi @MarcoViola88 , here is my pseudo abstract:

Recent research suggests that the uncanny valley hypothesis may constitute an insightful framework for explaining negative social inferences from faces. One potential explanation is that stimuli that are difficult to categorize may evoke a negative emotional response. Do androgynous faces follow a similar pattern? Indeed, some studies showed that the challenge of categorizing androgynous faces into binary sex categories serves as a metacognitive factor that can contribute to negative social judgments. We conducted cross-cultural research to explore how categorical uncertainty influences social judgments of androgynous faces. Method: Categorical uncertainty was measured using reaction times in the Face Evaluation Task. Trustworthiness, creepiness, and perceived shared moral values were measured using Likert scales. In Study 1 (Spanish-speaking sample, N = 76), androgynous faces were rated as more trustworthy, less creepy, and more morally similar compared to sex-typical faces. Although androgynous faces were more difficult to classify into a binary sex category (female vs. male) than typical-sex faces, this cognitive difficulty did not affect the social judgments of the faces. Similar results were found in Study 2 with an Italian sample (N = 45). Study 3 (Spanish-speaking sample, N = 85) revealed an overall positive bias towards androgynous faces, particularly compared to male targets, even after accounting for morphing procedures in stimuli selection. This research indicates that under certain circumstances, a positive social bias towards androgynous faces can exist independently of categorical uncertainty.

[marcosperduti]

Here there is a proposition for our (me and @DominiqueMakowski) abstract:

Emotional response toward fiction and the underlying cognitive mechanisms

In the philosophical literature there has been a long-lasting debate on the nature of emotion toward fictional characters and events. Neuroscientific studies have underlined that processing of fictional entities, compared to real ones, engages to a lesser extent midline cortical structures that are involved in self-referential processing. Given that appraisal of self-relevance is a key mechanism in leading to a full-blown emotional response, it is likely that emotional response toward fiction would be dampened. On the other hand, processing fiction engages fronto-parietal regions involved in executive processes. Interestingly, it has been shown that executive functions underpin emotion regulation. In a series of studies we tested the hypotheses that appraising a stimulus as fictional would downregulate emotional reaction, and that this effect would be modulated by self-referential processes and interindividual variability in executive functions. In the first study (29 participants), we presented short emotional (negative and positive) and neutral video as fictional or real. For negative material, we found that subjective emotional experience (valance and intensity), but not physiological arousal, was reduced in the fictional condition. Moreover, the amount of personal memories linked to the scenes, independently of their nature (negative and positive) boosted the subjective emotional response. In the second study (37 participants), we aimed at conceptually replicating our previous results using different material, and at investigating the role of executive function in fictional reappraisal. We presented participants with negatively valanced pictures of varying emotional intensity preceded by short texts describing them as either fictional or real, and assessed participants’ executive functions (updating, switching, and inhibition). According to the previous study, we observed a diminished self-reported emotional arousal when pictures were described as fictional for high- and mild-intensity material, but not for neutral material. While no effect was observed for physiological arousal. Moreover, the amount of down-regulation in the fictional condition was predicted by interindividual variability in updating performances. The third study (33 participants), aimed at replicating our previous results, and elucidating the neural (EEG) and physiological correlates of this effect. We confirmed that stimuli presented as fictional, compared with those presented as real, were subjectively appraised as less intense and less negative. Moreover, they elicited lower physiological reaction (skin conductance, heart-rate deceleration), and lower neural response (late positive potential). Self-referential processes complexly modulated subjective and neurophysiological measures of the emotional response. Taken together, these findings show that appraising a stimulus as fictional can diminish the emotional response, and that this effect could vary depending on the valance of the stimulus (positive, negative), interindividual differences in higher cognitive functions, and personal relevance of the stimulus. We will discuss the relevance of our findings for the current debate on the negative bias toward AI-generated material.

[marcosperduti]

@MarcoViola88 I can reduce the length if necessary

[MarcoViola88]

Hi everybody,
a co-author of "Real is the new sexy", Alessandro Demichelis, volunteered to present it at the symposium. Abstract at the bottom of this message.
That brings the # of speakers to 3 (Olivera-La Rosa, Sperduti, Demichelis). We may have time for a 4th slot: how should we fill it? I have a few options in mind:

1. Rather than having a 4th speaker, we just let @marcosperduti discuss all the studies -- they are a few, and may require a bit of unpacking.
2. We (most likely @DominiqueMakowski) present the preliminary results of the current study.
3. We try to invite someone people who can say something relevant for the current purpose. If you don't mind a philosopher commenting on empirical result, maybe we can invite M argherita Arcangeli, Marco S. and Dominique's former co-author?
4. Other option: Liefoghe, who authored this paper caliming that faces labeled as artificial are trusted less https://online.ucpress.edu/collabra/article/9/1/73066/195773/Are-Natural-Faces-Merely-Labelled-as-Artificial
5. Other option: someone working on the neighbouring topic of AI-generated art ? I may know a few people, but let me know.
6. Open to your proposals!

My preferences in order: 3 > 1 > 4 > 2 (pending availability of the person who should do the talk of course!) -- let me know yours ;-)

@marcosperduti unless we pick option 1, would you mind reducing the lenght of the abstract a bit? Down to 300 words should suffice!

The proliferation of artificially-generated images indistinguishable from real pictures impose several societal challenges Extant studies suggest that believing that faces are or can be fake decrease their trustworthiness, whereas allegedly fake non-sexual videoclips are less arousing than real ones. Thus, to test whether the “purportedly unreal = less arousing” effect also applies to sexual contents, we performed 2 pre-registered online studies. In study 1, participants (n=57) saw 60 images of male or female models in underwear. They were told that pictures could be either genuine photos or AI-generated (unbeknownst to them, all stimuli were real). For each stimulus, they were asked both whether the stimulus was real and how much they were sexually aroused by that image (on 6-points likert scales). In study 2, participants (n=108) were shown the same images, but this time they were presented with 30 real and 30 allegedly artificially generated pictures in two different blocks. They had to rate their sexual arousal (on a 6-points likert scale). We found that realness, be it self-rated (study 1) or experimentally manipulated (study 2), was a significant positive predictor of higher arousal. These findings have significant implications for understanding the impact of deepfakes and suggest a robust correlation between the assessment of authenticity and the potential for experiencing arousal.

[marcosperduti]

My preferences in order: 3 > 1 > 4 > 2 (pending availability of the person who should do the talk of course!) -- let me know yours ;-)

I think that having someone (a philosopher) commenting our results would be definitively very interesting. It would be great to have Margherita, since she knows very well our works.

I will reduce my abstract and repost it as soon as possible.

[marcosperduti]

Here there is my shortened abstract (256 words):

In philosophy there has been a long-lasting debate on the nature of emotion toward fictional characters and events. Neuroscientific studies have underlined that processing of fictional entities reduces the recruitment of cortical regions involved in self-referential processing, and boost activity in fronto-parietal regions involved in executive processes. In a series of studies we tested the hypotheses that appraising a stimulus as fictional would downregulate emotional reaction, and that this effect would be modulated by self-referential processes and interindividual variability in executive functions. In all the studies we presented emotional laden material - videos (Study 1, 29 participants) or pictures (Study 2, 37 participants; Study 3, 33 participants) – as either fictional or real. We collected subjective, physiological and neuronal data characterizing participants’ emotional reaction. In all studies, we reported that negative stimuli presented as fictional were subjectively appraised as less intense and less negative. Moreover, we reported that they elicited lower physiological (skin conductance, heart-rate deceleration), and neural response (late positive potential). We also showed that self-referential processes complexly modulated the effect of fictional appraisal. Finally, we reported that the amount of down-regulation in the fictional condition was predicted by interindividual variability in updating performances. Taken together, these findings show that appraising a stimulus as fictional can diminish the emotional response, and that this effect could vary depending on the valance of the stimulus (positive, negative), interindividual differences in higher cognitive functions, and personal relevance of the stimulus. We will discuss the relevance of our findings for the current debate on the negative bias toward AI-generated material.

[AleAnsani]

Hey guys, if you're ok with that, I could fill the 4th slot. As some of you know, I've just started to collect the data for an experiment on the role of the human being (vs. purported AI) in musical performance. The main aim is to investigate the extent to which the mere presence of a human being modifies the aesthetic judgements / emotional impact of a musical performance. I might have some preliminary data by the symposium deadline ;)
How does that sound?

[MarcoViola88]

Ok, as soon as @AleAnsani send his abstract wrap everything up and post here for final approval!
I look forward to it :-D
The organisers told me symposia will have 80-100 minutes, so I guess we won't have time to also present the results of our common soon-to-be-published-in-Nature project, right?