INTRODUCTION
The past year alone has shaken the core of what ‘AI’ represents to the public. With the rapid popularisation of tools such as ChatGPT, Dall-E and Midjourney, AI has now become a bridge between people and the skills they previously did not explore. It is rapidly growing into an extension of how we think, a third arm that exists in the Digital Ether.
This expansion has come with its fair share of anxieties, such as AI gaining sentience for the worse and the general loss of perceived human value. Through this inquiry, I want to explore how AI can help amplify a core element of the human condition: that of empathy. This is fuelled by the observation that in many ways the growth trajectory of AI is evolving to reflect the innermost human need for connection.
Since AI was introduced to the public, it has acted as a language tool between us and the digital ether. An early example of this is the ELIZA program, developed in 1966, it was a natural language processing program that simulated a conversation between a therapist and a patient. (Weizenbaum et al, 1966) It was able to respond to user inputs in a conversational manner. Fast forward to now, AI has become an increasingly prevalent part of our daily lives. We converse with chatbots and virtual assistants, personalised AI algorithms dictate our consumption patterns, and smart home devices start and end our day. Deeply influenced by Jungian philosophy, I believe this continual use of AI has fed and fuelled it to evolve into a digital manifestation of our Collective Unconscious.
A SHARED LANGUAGE
Carl Jung defined the collective unconscious as a shared, inherited reservoir of images and symbolic archetypes that are common to all human beings, regardless of culture or individual experience. According to Jung, the collective unconscious is composed of the deep, fundamental patterns of the human psyche, which are expressed through myths, symbols, and dreams. (Jung, 1912) With text-to-image tools such as Midjourney and Dall-E, we are now living in an era wherein we get to experience our Collective Unconscious, consciously and collectively. This leads to the deduction: the existence of a Digital Collective Unconscious implies the existence of a universal, shared language within the Digital Ether.
The possibilities of exploring this shared language are endless, and we are only beginning to scratch the surface. To delve deeper into this exploration, this article aims to merge the worlds of human empathy, AI capabilities, and the Digital Collective Unconscious through the following question:
Is it possible to craft a visual language, aided by AI, to enable individuals with hearing impairments to experience music at a large scale?
There are a few factors fuelling this question:
Music can be experienced in several ways, and for those hard of hearing, It can be experienced visually and with the help of vibrations. The hearing-impaired rely on their immediate surroundings to experience music, rather than a shared experience. There is thus a need for a collective language where these populations get to experience music together at a large scale.
The Digital Collective Unconscious, or the Digital Ether, may have already created visual definitions of terms. This inquiry wants to further investigate this hypothesis, and understand how the Ether defines our verbal language.
The combination of these factors might help us create a new way of experiencing our perceptions of the world, through a medium of our own making.
DESIGNING A LANGUAGE
A. Context:
To contextualise this inquiry, the author has chosen the medium of album art. A song and its album art go hand in hand, and although our target audience might not get to listen to the music, they can shape their perception based on the visual elements of the album art.
In addition to the primary album art, the author proposes individual ‘song art’, where each song has a separate artwork that visually communicates the essence of the song. These songs can then perhaps be experienced more holistically with haptic devices.
B. Method:
To carry this out, AI tools ChatGPT and Dall-E have been used. ChatGPT was used to describe each song visually. These descriptions were then input into Dall-E to create custom song art covers.
C. Process:
Testing the waters
To begin with, I wanted to understand how the software visually defines key components of music. Key elements such as ‘Rhythm’, ‘Harmony’ and ‘Pitch’ were tested first.
a. Results for Rhythm:
Upon further exploration:
b. Results for Pitch:
On probing further:
c. Results for Harmony:
Round two:
Through this short test, we see that Dall-E successfully interpreted these aspects of music through a flow of colours and forms. It appears to have an existing understanding of visual definitions of sonic terms. With further variations, we may be able to create a standardised understanding of visual language within music.
Prototyping Kings and Limbs
The album ‘The King of Limbs’ by Radiohead, released in 2011 was chosen for this part of the experiment. Radiohead, a cult favourite, is famous for its experimental sounds and vulnerable lyrics. Their album art, consistently designed by the artist Dan Rickwood (Stanley Donwood), has been synonymous with the identity of the band. King of Limbs in particular left a lasting impact on the band's discography, as it marked the beginning of a new era within their musical experimentation.
Above: TOP (L-R): The King of Limbs, 2011; In Rainbows, 2007; Amnesiac, 2001; MID (L-R): OK Computer, 1997; A Moon Shaped Pool, 2016; Hail to the Thief, 2003; BOTTOM (L-R): Pablo Honey, 1993; Kid A, 2000; The Bends, 1995.
Each song was analysed by ChatGPT separately before being input into Dall-E. It was asked to visually describe what the song feels like. The following is a table with ChatGPT’s description of the song, parts of this description were then fed into Dall-E's prompts for visual interpretation.
Key Learnings:
There is, indeed, a Digital Unconscious, a byproduct of our own making. It is the sum of all the inputs we ever allowed into the digital realm.
Both mediums, ChatGPT and Dall E, effectively understood the purpose of the prompts, and were a good medium for expressing the intangible.
Dall-E had surprisingly large room for the abstract, and did a good job of translating feelings into visuals.
Viewer discretion plays a big role in the selection of the artwork that best suits the song. Therefore the ‘best’ artwork applicable is open to interpretation.
Each artwork does a good job of conveying the tone of the song, however there is scope for improvement in bringing a sense of visual continuity across them.
The lack of continuity hinders the process of calling this the creation of a language per se, however it does open up possibilities of investigation into AI’s existing understanding of the abstract, which in turn can be used to create a standardised language.
Probable use-cases:
Album art has the opportunity to expand exponentially, reaching out to not only those who are hard of hearing but also to the general public.
Although it has been designed for a specific target audience, it creates a platform for a new shared experience of music, one that is rooted in the intangible.
It creates new opportunities to explore album art as a medium of communication for artists. It can provide them with added flexibility, as well as new mediums of connecting with their listeners.
CONCLUSION
I believe that AI can act as a mediator between us and our need for understanding the intangible. The ‘Digital Ether’, an object of our own making, offers us a new perspective on how we view the abstract, and creates opportunities to use it as a shared language. The act of cocreation between us and AI can be a seamless experience, and there is much to gain from recognising that human empathy is indispensable. It will continue to be the secret ingredient that generates authenticity, and creates a sense of connection among the products and services we design for eachother.
Comments