Computer vision and the arts

Friday March 3rd

As we draw closer to 𝜋 day, our team is starting to lock in the various activities and events that we envisioned, I had the great opportunity to interview Dr. Jiamin Bai, a research computer scientist with over 15 years of experience in computer vision. We discussed the intricate connections between mathematics and the arts, and how this relationship has evolved in light of recent developments. I would like to extend my thanks to him for his insight into the cutting-edge research and for agreeing to address this year's 𝜋 day event at UWCSEA East, Singapore.

Due to emerging mathematical descriptions of our reality, computers have gained the ability to describe many phenomena, previously understood only through human consciousness, and perform computation on this understanding. Note that this in and of itself does not constitute the consciousness of computers, but rather the necessary preprocessing of raw data using mathematical models to enable the manipulation of previously abstract definitions of reality. As humans, we are already intimately familiar with these descriptions and our surroundings as result of our brain's inherent data processing capabilities which we've evolved to survive. 

Computers are getting better at gaining an understanding on what things are and how to exploit this understanding, but not how to formulate original ideas about them - a subtle but significant difference. With these breakthroughs, it has become increasingly evident that the next paradigm of AI and robotics is upon us. Of course these changes, as with all overarching changes, raise some interesting concerns for the future. Concerns regarding the nature of our knowledge and how it can be expressed; the core of creativity and the arts; and the justification of mathematicians, computer scientists, and technology-enthusiasts to impede the practice of artists who have spent years honing their craft. 

On a personal note, I believe that these advances will be crucial towards the world we build in the future and are justified in the march of progress. That isn't to condone misuse of technology or replacement of several jobs in society but rather as a note of the natural evolution in the most fundamental existence of humanity. As we move away from industrial and manufacturing processes, we will gain access to more specialised and efficient structures. These general models of our reality will be able to fit a plethora of needs and sizes, and accommodate invidualized commercial and social needs alike. In the future, we won't buy a book which contains fixed content, rather we will buy a more fundamental representation of the textual media which the author is trying to portray. The media would thus imbibe a core idea or universe in which an an infinite number of stories could be generated. The author would act as a guide to design a formula for an experience or an emotion. Everyone's reading experience would be different and tailored around their own fantasies to best appeal to their preferences and unique identity. Instead of buying a copy of Lord of the Rings you would instead buy Middle Earth in the style of Tolkien. Instead of buying a set of clothes, we will instead buy a "mood" of clothing that will be custom fit to any size and shape, and still reflect the basic statistical truths of the style the fashion designer wanted to express. Buildings could be designed to merge seamlessly, interpolating the wants of the future tenants.

I'd argue that there is a creativity in mathematics parallel to that of artists, and recent discoveries have enabled mathematicians to take data and objects, with all their endless properties, and express them statistically. This creative desire to express an idea or an object or an emotion is something which I think is transferable between artists of all kinds as well as mathematicians and scientists. The idea of beauty can be seen referenced in all subjects and studies, from mathematics to the arts - it's all a matter of perspective.

AI generated rendition of elephants in different styles of art

Additional Resources:
- How computers are learning to be creative - Blaise Agüera y Arcas