This week I focused on refining my interests and approaches into actionable ideas. I started with several topics such as Neural Aesthetic, interface design and metaphysical space and connected them to tools that represented these topics. I chose to focus on two potential designs instead of three as these two ideas stood out as much more developed than the rest and they each tie together several of my interests and technological capabilities.

 

AlmostOS

This project springs from my interest in interfaces and interface design, specifically defunct operating systems. The history of interface development has been guided by western design principles, focused on simplification, usability and limiting user decision-making. Tracing back the history of interfaces I found many alternate approaches to user-machine interaction, going back to the first analog computer Antikythera which included instructions on the side and cranks to work the astronomy tracking mechanism. Science fiction writers like E.M. Forster imagine future interfaces as complex and tangible; “The Machine Stops” imagines a future of living inside a supercomputer controlled by hundreds of buttons and knobs that each indicate a different task or request. Early computers used physical interfaces and interactions, such as teleprinters and punch cards used to feed commands into the computer. Command line interfaces allowed users to interact with the computer in real-time, but the agency was all on the user as the interface provided no prompts, choices or suggestions. After the widescale use of video displays, operating systems with a focus on accessibility and user function created the 2D (now 3D) virtual space inside computers that guided users into interactions. Along the way, compatibility and large technology companies steered development, and the many possibilities of interacting with computers and displaying information were narrowed. One example is TempleOS, a fascinating operating system first released in 2003, coded entirely by Terry Davis, a troubled computer scientist who believed his operating system was the Biblical third temple. The very fact that he created his own coding language Holy C as well as an entire operating system shows a loss of capability and flexibility in the development of interfaces, and the very idea of switching operating systems on a computer today is daunting and nearly impossible. Davis reimagined an OS that put user creation first, designed for creating games and animated graphics as well as cryptic “messages from God.”

I would love to explore this idea of defunct operating systems and user interfaces that prioritize creation and control over ease and accessibility. I want to create a website in the style of an old operating system, based on Macintosh 128k, Windows 1.0 and TempleOS (and possibly others like SunOS and neXTSTEPS). The website will be an interactive experience that mimics the functionality of an imagined operating system, or perhaps even a speculative design of an alternate timeline where interfaces were designed under different values and systems.

 

 

AI Nightmare

As part of my interest in the neural aesthetic and AI generated images, I would love to continue my exploration of VR experiences through AI’s interpretation of nightmares. GAN programs such as Big Sleep and Deep Daze use text to image generation, creating an image from a few words or sentences based on huge databases of image and associated text. While some programs are focused on photorealism and object identification, Big Sleep and more so Deep Daze offer a more abstract interpretation of text. “Nightmare,” in combination with other words and styles produces fascinating results, and begs the question, what is a nightmare to AI?

I would love to apply this to a 360 immersive space with positional AI generated sound. I’m inspired by VR artists such as Sutu and AI artists like Memo Akten who create immersive, abstract pieces rooted in human experiences. I want to create a VR journey through an AI nightmare by training models and generating “nightmare” videos and converting them to 360 using After Effects. I’m also considering incorporating interaction, whether it’s through using webcam and audio feeds to bring the user into the experience or offer an insight into the decisions AI makes when generating content.

 

I’m still not sure which direction to go, but I do love the idea of working on a machine learning project for my thesis as I’m currently building my ML portfolio. I want to create an experience that’s still accessible and usable, such as a website or web VR, but immerses users into an unfamiliar world with unfamiliar dimensions and interactions. I’m excited to start working on my thesis and see how it develops over time!