After doing more indepth research on GANS and using my dataset of photos to create new imagery, I realised the output I wanted wasn’t going to be viable in this time frame and skill I have. I saw the labour required to create a GAN through Anna Ridler and her Tulip project in which she had to photograph over 10,000 tulips to create her imagery. I began to think more widely about image making in a digital, post-ai context. There are some softwares where you can trial having server space and the software infrastructure to begin formulating a GAN such as Microsoft Azure, however, I realised the monotonous, work I’d have to do and wanted to challenge myself within graphic design. There is the idea that the data-set and algorithm are part of the creativity. But to explore this, I felt I needed some basic vocabulary in this field, which I don’t currently have but intend to learn later on. To formulate my image making process I explored other ways of incorporating AI technologies in my work. I experimented with AI Image upscalers and programmes such as deep dream. I was interested to see how various technologies abstracted or took away from the original image. These processes can be used in tandem and overlapped: (deep dream – colour photoshop – reduce colours – pixelate) to extend the abstraction. These are some of the resources I used:
https://www.freecodecamp.org/news/how-ai-can-learn-to-generate-pictures-of-cats-ba692cb6eae4/
https://teachablemachine.withgoogle.com/train
https://deepdreamgenerator.com/
https://azure.microsoft.com/en-us/services/machine-learning/
https://www.nvidia.com/en-us/research/ai-playground/
Topaz giga pixel: https://www.youtube.com/watch?v=rl1pbEG2kgA
Below is a digital copy of my visual experiments that I printed to display in my pin-up:
I thought about how typography could interact with these images. I think it’s interesting to reflect my image making process in my narrative. I’ve enjoyed exploring the open nature of ai generated text and often come back to the web browser AI software https://talktotransformer.com/. I just find the results so interesting and how it picks up the style of the previous language given to it especially within the context of the fantasy world inherent in text-based adventure games. I’ve also explored how softwares such as https://botnik.org/ can effect text. I’ve also had the idea to pair my imagery with the text of Walter Benjamin. I previously explored his text “art in the age of mechanical reproduction” and it inferred my understanding of the implication of the digital in a really interesting way. Similar to Sapere Aude I may work with a text to infer my graphic lanugage: I thuoght about researching Benjamin’s essay “The Task of the Translator”.
“http://users.clas.ufl.edu/burt/deconstructionandnewmediatheory/walterbenjamintasktranslator.pdf“
Interestingly, in my research I came across a interactive piece called “AI Dungeon”. It felt odd to see because it was my combination of thinking points but executed in a very technical way. It’s fun to play around with but it didn’t really stick with me. It is constantly generating new paths to follow but you can say and do anything, it’s a completely open system. I feel as a user you are given too much choice and free will. It’s odd but interesting to critique despite the ground breaking technological feat. It’s like an extension of talk to transformer but infinitely sprawling and contextualised in a classic text-based adventure fantasy world. In my mind there has to be an equal relationship between something that is inferred by man and machine. In my opinion, with the absolute freedom it allows, the machine runs wild and overpowers the experience.
I had a brief think about the font’s used in the old text-based adventure games – how the less than perfect typographic communication was inferred by the technological limitation of the computers at the time. I thought about how I could combine all these aspects into a digital interaction. Through research, I realised the best way to create text-based adventure games is to use a software called twine.
Within this process I’m also thinking about narratives and how stories are told. I am beginning to think about how texts display narrative in different, nontraditional ways. I have thought about some books such as “Beowulf”, “ulysses”, “Petit Prince” and “On the Road”.
Cameron Aitken Graphic Design 