The aim here was to create 6 quick prototypes around the design questions that I had in 6 days, each day starting on a new prototype and ending on one. The thought was to create 2 research prototypes, 2 technical ones and 2 design prototypes with the aim of exploring the design questions from each direction and narrowing down towards a few ideas to explore in detail.
Topics explored: The future of Social Interactions & Reconstructing History through MEMEs
DAY 1 # Research Prototype - Interview with an Expert
I interviewed an AR Designer working in Oculus Research in Facebook. My aim was to understand what kind of research work is going on in a leading tech facility in this domain. My questions were focused towards understanding the kind of experimental projects that they work on as well as any insights in marketing trends that their research has shown.
Key points from interview
“The biggest mistake that people do is think of MR as just a display medium, when it’s much more.”
Combine AR with other tangible interfaces to create a product with holistic experience.
Think of what would be the best way for users to complete a particular interaction
Think of specific and unique use cases where MR could be helpful
DAY 2 # Conceptual Prototype - A Comic strip on Genuine Intelligence
The concept of Machines learning about humans and Genuine Intelligence could be an interesting direction to built speculative concepts on
DAY # 3 Research Prototype - Reading the book Memes in Digital Culture by Limor Shifman
Key quotes from the book:
Like many Web 2.0 applications, memes diffuse from person to person, but shape and reflect general social mindsets.
Internet memes play a key role in contemporary formulations of political participation and cultural globalization.
At any moment, many memes compete for the attention of hosts; but only ones suited to their sociocultural environment spread successfully, while others become extinct
Read more research on impact of memes in socio-political and global events.
Use the current research findings as baseline for storytelling.
DAY # 4 Conceptual Prototype - A Voice based Social Media
Designed a concept for a voice based social media platform.
Think of how it can be combined with tangible/gestural feedback
Design a platform that uses multiple inputs to complete a task
DAY # 5 Technical Prototype - Reconstructing History by ML
Continued on my previous meme-based prototype by training an ML model on archaeology book and journals and then fed it texts from memes.
Refine the model by cleaning up dataset.
Try including some relevant data about Sci-fi
DAY # 6 Technical Prototype - Feature extraction from memes
Wanted to create a text and feature extractor from image so that when I feed the ML model an image, it could extract the text and use it as a seed to generate content. Tried out different platforms for this like ML5 and RunwayML but figured that they have certain limitations that did not allow this. Finally started working with the cloud vision API of Google but did not have enough time to complete in a day. So used an already functioning application of the vision API (Google Lens) to extract text from images.
Runway ML (x)
Cloud Vision API
Complete this using Cloud Vision API
Extract labels and landmarks from memes and experiment with that