Unit 3:
Tertulia
On October 19th 2023, I was invited to speak about my research by Nele Bergmans, a fellow MA Fine Art student at Camberwell, in the lunch salon event known as ‘Tertulia’. This event sought to create a relaxed environment of discussion and the sharing of research between MA Fine Art students. I spoke in my own presentation about the theme of AI and animism, the work of Jake Elwes and the Stable Diffusion platform I have used to render my videos. I found this experience valuable, not only because I feel it is important to discuss and unpack AI technologies through art, but also because it was exciting to see how many students feel passionately about the ongoing climate surrounding AI, and to hear what they had to say.
I believe artists especially must partake in the conversation surrounding AI, since AI threatens the livelihoods of artists through harvesting their works in datasets. The students expressed the importance of open mindedness in conversation and across artistic practices and pathways. Though some admitted they were against AI art at the beginning of the course, they had since seen uses of AI in artistic practice that had changed the way they viewed the technology and its potential. The inability to reverse time on the development of AI was also discussed with some acceptance; that perhaps using AI art with full acknowledgement of the process and nature of the technology is the best way for artists to engage with the field of this technology that doesn’t appear to be leaving our daily lives any time soon.
Maria B Brito from the MA Fine Art Photography pathway also discussed her own work and research during the same session. I noticed some parallels between our interests: including the interest of where the definition of a person/figure/identity is dissolved, as it grows more abstracted in her case, or more enveloped into the generation of an AI model, in my case. I suggested that we should do an exhibition together after the course submission deadline.
Instaweek
After the start of Unit 3, until the run up to the Summer Exhibition, each of the Computational Arts Pathway students had the chance to share their work and research in instagram stories for one week on the MA Fine Art Computational Arts Instagram page. I made the following posts which were shared on the @ma_faca Instagram page and that can still be found in the story highlights under my own instagram name: @jessicabraunerartist, here:
https://www.instagram.com/ma_faca/ [Accessed 27th October 2023].
Although these videos primarily explain my Unit 2 work and research, there are still glimpses of my Unit 3 work and many of my ideas from Unit 2 have progressed over the course of Unit 3.
I found that the students from the new MA computational arts pathway in the year below me knew who I was before we had been introduced, which made me wonder if these videos had reached them. I hope so, as the computational arts pathway is a new fine art pathway on the course, and I feel it has room for expansion and for more students to get involved.
All of the information discussed in these Instaweek videos can be found either in the Unit 3 sections of this website, or in the Unit 2 pages on this website:
https://jessicabrauner.myblog.arts.ac.uk/final-artistic-outcomes/2/
https://jessicabrauner.myblog.arts.ac.uk/artist-statement/2/
https://jessicabrauner.myblog.arts.ac.uk/current-art-practice/4/
https://jessicabrauner.myblog.arts.ac.uk/critical-reflection/2/
A Revision of the Archive Mini Project after viewing Eryk Salvaggio’s : ‘Flowers Blooming Backward Into Noise’ 2023
In the Archive Mini Project of Unit 2, I asked: ‘is LAION-5B in itself an archive?’. After watching Eryk Salvaggio’s manifesto work online: ‘Flowers Blooming Backward into Noise’ 2023, I have considered AI datasets once again in relation to the concept of the archive. AI generated images are noted in Salvaggio’s work as an ‘infographic’ about the dataset (Salvaggio, 2023). The datasets for AI and for Archives can both hold biases but while data in archives can be more explicitly contextualised and categorised according to date, place and source, the data that is used for AI image generation is categorised only according to the image labels and the correlations, or pathways, between image compositions.
Salvaggio’s work describes the dangers of categorisation in this way through mentioning composite photographs – photographs that are made through blending similar photographs into one image. The video describes how Francis Galton used collections of photographs to ‘create a common face amongst prisoners.'(Salvaggio, 2023), in order to suggest who would be more likely to commit a crime based on their facial features. Francis Galton was also mentioned as a figure who originated the eugenics movement in the Genetic Automata exhibition I saw, on my contexts page. The risks of lacking careful reflection and critical thinking when observing AI generated images of human beings from biased datasets is evidenced through this connection. If archives placed their image data into the categorisation of similar images according to composition, or tones without the context behind the image, or combined those images together, the archive would be left with visually stereotyping material as a tool for filing information. This would only perpetuate prejudices and further manipulate the information that is left from collective histories.
In seeing the connections between these discussions surrounding AI in my research, I should like to deconstruct this further in my work, going forward. In using AI, I must remain critically conscious of the dataset biases and how AI images are generated. I am left considering how I can highlight the dangers of these patriarchal and colonial biases within AI datasets further in my work. How might these biases be challenged when using AI going forward? Though I have already explored some of these themes in combining myself bodily with the AI generated feminine mythical figures of sea nymphs and harpies, this was more closely tied to the fiction surrounding Shakespeare and the spiritual connotations AI can inspire. In the future, I might seek to focus more closely on my own developing narratives, as well as on how the stereotyping methods of AI generation can be challenged more explicitly.
Bibliography:
Eryk Salvaggio. (2023). ‘Flowers Blooming Backward Into Noise (2023)’ Youtube. [online]. Available from: https://www.youtube.com/watch?v=zNA7sPm-zlQ [Accessed: 26th October].