this post is a work in progress
What are the possibilities of thinking through or with a machine learning process working with datasets of spoken English to train a recurrent neural network (RNN) trained on sound and music to explore the complexity of the depth of being conscious and unconscious?
La la la
Like machine learning processes, to go ahead with this research project, I must begin with the already here and much that is happening always slightly ahead of the instant.
In the beginning, i thought this project would be about machine learning neural networks and language. From there, I have rambled into discussions about the brain(s), emotions, human bodies, embodiment, the inexpressible, the in unknowable, and the un incomputable. Towards the end of this entry is some background to some works that in part led to this project.
So, to the just about
raw language sponge etc ((((((((((draft thought How machine learning processes reveal the importance of the rawness of expression and feeling prior to a human learning a structured language. This is what lacan relates a lalangue. Unconscious and conscious and the relationship of intelligence imagination etc as part of the whole.)))
I am following the non-sense
Toward the unfolding from hyperspace
This is a slice of imagining, a very close look at how I think feeling thoughts is material, the sensation of vastness. It is also a way I can be in a world that is there when I close my eyes. I start to wonder if a slice of hyperspace, twelve dimensions, might be similar if a part of it was sliced away and held on a microscope slide. Is experience of the slicing, pinching, and restitching of parts of the multidimensional to make it possible to be in it to comprehend the material and more than I don't have the vocabulary for of the universe with the limitations put on the human hardware? In such temporal bodies can computational power free time and space to make it possible to activate, rewire the human hardware to experience much more of that around and expand embodiment.
back to the background
Some time ago—it feels like a long time—I wanted to express my thoughts before they were mediated by language. I have come to relate to this drive as wanting to speak a dream. This want is so strong and backed up by effort, but despite this, I find it really difficult to do.
The notion of expression of the rawness of thought, its moment in an instant, is a constant in my work; it remains difficult to find ways to know if thought as raw material is ever possible to express.
Awaking to re-memberable dreams is quite something. So fragile is the moment of hovering between sleep and awake that disruption to the space between is, I think, a fleeting duality of unconsciousness and consciousness slipping space and place until either sleep or awake state is returned to.
When I find myself in conscious moments of waking up, I am aware of a need for comprehension of the feelings resonating from the dreams that remain. Those feelings start to resonate elsewhere as the words to express them are desperately searched for. This is when I immediately sense feelings becoming something else in their translation into meaning.
My brain-activated imagining translated how to speak this dream, which is driven by a chronological recolocation of the anything-but-linearity time of dreams.
Thinking about the notion of a deep object in relation to language guides the new work I have made during this residency and research, but importantly, I have, as I describe a little above, been able to frame my previous efforts to work with language. This text will reflect on three pieces that already existed and have reverberated into several other pieces, which I will also discuss and conclude with the work I have yet to make but have started during this process.
In 2016, I began to think seriously about my difficulties with writing, reading, and expression, especially using my voice and how I might move them into a place where they, it wouldn't hold me away from doing things I wanted to.
Eventually I began some work around the alphabet, which underpins many of my difficulties. The alphabet is quite an extraordinary component of a machine language: 26 letters from a whole as an abstract linear set constantly reassembled further abstracting its linearity.
It began with confusion about how I still found it tricky to recite parts of the alphabet. I could only know the location of one letter in relation to another by reciting the entire thing from the start. The alphabet, the a, b,c, remains elusive to recall. So I began with simplicity, recording myself repeatedly reciting the alphabet as insurance at times to see how long I could keep going.
Work 1 Alpha-beta, details
Y,z
Utter[us]
Language
deep object
affect
Since I began working seriously with machine learning (2019), it is not an overstatement to say there has been an explosion in the presence of this technology.
My intention, which I set very clearly, was to keep far from NLP as I believed then and even more now, that the construct of nature, humans, and language is an illusion. I work with machine learning locally, in sound I only use my datasets, as regards images; I'm also trying to work locally, but the cross over with other models is something I want to avoid.
As much as I say language, it is now not just about that language, the structured ordered learnable one; it is also about the unique rawness of an expression, lalangue, ---
Work 2 Utter_[us], details
The experience
Raw data and the expansion of the human; embodiment and beyond. Topology and materiality into the realm of scientific reductiveness predictive modelling.