<< Diffused Referencing >>


Diffusion Models are a digital technology allowing for graphic output, triggered by (human) prompting of the tech, with a well-groomed (human-written) textual phrasing. This here is a verbal play on these models, back to human-struggled and confused text. It’s all hand made. If you are an engineer or scientist you might find it irritating or infantile. Yes, it is human play. Hence you are most invited reading it.

Please read with as much “wink” you wish to add, to make it digestible.  [ diffused translation: imagine a solid, colorful Polyoxybenzylmethylenglycolanhydride wink-diffusion slider on a meaning-making mix console]

as for the above visual: resist being mesmerized. (A diffused paraphrasing from an utterance by an actual human academic)

Welcome to the real of Referencing Technology (RefTech). [diffused translation: welcome to the extremely large models with billions of features and a referenced database, as part of a pattern recognizing technology]. There is meaning here, you do think to recognize something. Your brain is obsessively trying to place it.  

We trust the referencing automated by the RefTech to output a nonreferencable. In the nonreferencable we trust. [ diffused translation: imagine a visualization of an unreachable heavenly ivory tower, domesticated by a wizard behind a curtain telling you: “it’s all your creativity, kid!”]

Some might imagine that for thousands of years we have been preparing for our RefTech by creating narratives and spin-off institutions of that what should not be named (ie not referenced) and by diluting that what could have been named (ie observed, tested, deduced, inferred or otherwise transparently thought-exercised) with gradual, unveiling, and yes, humanly failing effort. [diffused translation: envision a visual of an organic, mysterious yet liberating RefTech as the 3D rendered misinformed cherry on a simulated cake, presented to celebrate humanity’s greatest achievement: your tweaking and prompting into the world]

The recipe is simple: forget roots and assign metaphysical magic to that what is no longer known, not yet known, or made to be kept unknown. [diffused translation: the unknown known of a photorealistic word to pixel, birthed by your hyperrealistic yet de-textured mind to smoothened hand]. 

What better augmentation to this non-linear development by creating not only that what should not be referenced and also cannot be referenced —by giving users a virtual slider or two to fiddle with— and out comes an uncanny close-enough. 

In a desert of the real can grow a beckoning flower that sirens ownership. Is this desert from, of or via Boudrillard?

Were his words diffused, resynthesized and co-opted by the first film in the Matrix series & then later taken by Žižek for a book title or did the imperial uber-randomness strike twice? 

Or trice: Boudrillard himself might have reformatted the idea of the desert from other authors who described “real virtuality,” hyperreality, “death of the real” or deserts as ideal places for forty days long visions of real proportions.

These (non)references could understandably be upsetting to some reader (eg if the ontology of one’s measurable and almost-tangible deductive logic becomes endangered or upset), or too hyped by yet others who “enjoy” the less measurable. [diffused translation: false dichotomies are such easy and low hanging fruits]. 

RefTech as a “wished for” technology to allow more anthropomorphization. This offered by tactile interaction, into the idealized & utopian “geniusness” of creative output easily diffusible by the click of a “button.” Buttons and diffusions allowing it and oneself to become as-if-one’s-own. The “beauty” of the paradox does not go unnoticed realizing the single spacetimed button leading to (entropic) diffusion; thermodynamically atomic.

This process is fashioned all the while we are allowed to mechanize, simplify or deny the human, and the humane of the creative relationships, weavable within the real of you and me, ever slightly touching indexes. (Yes, in metaphorical counterpoint, some might now be reminded of Bob Dylan and his diffusions of man across the fields upon which winds howl. No, no one would think of a chapel here. The latter is the elephant asked not to think about and this so not to slide toward yielding it with historic and spiritual meaning).

Granted, with the RefTech we are offering and answering unknown-needs in a metaworld of unknown-urgency: the (lesser/greater) geniuses among us do gain access to tools for diffused referencing, while the mean of the common among us gain and repurpose play and distractors 

RefTech doesn’t enable the creator and  “prompt engineer”to reference all the referenced works used to create “real” & more-than-human output, as much as any word written here was once in a past referencable to its ancient Creators, yet now diluted and lost in the diffusions of human-mounted noise.

Welcome to the real unreal engines where history is ravaged; welcome to the pixel of a word spread out as a dust spec(k) of ridiculed substance into an unlabeled future.

references.


Sohl-Dickstein, J. (2015). “Deep Unsupervised Learning using Nonequilibrium Thermodynamics” https://arxiv.org/pdf/1503.03585.pdf

ImageGen by Google: https://imagen.research.google

Dall-E2 by OpenAI: https://openai.com/dall-e-2/

Huggingface: https://huggingface.co/blog/ and: https://huggingface.co/blog/stable_diffusion

crAIyon:https://www.craiyon.com/

StabilityAI: https://beta.dreamstudio.ai/home

midjourney: https://www.midjourney.com/home/

Automated prompt generation: https://typestitch.com

no I did not add all authors whom I diffused across and in between the above words.

—-•
questions:

• What could ‘RefTech’ be imagined to be(come)?
• What could “referencing humans” (be imagined to) mean to you?

source: (wink)


<< Wise intelligence-wise >>

While many of us strive for those around us to be less complete and complex humans,

we equally strive for our artificial output to be more anthropomorphic and life-like.

Is the place where both processes will find their nexus the singularity and a desert of the real we should yearn for?

—-animasuri’22

—-•
contexts:

https://garymarcus.substack.com/p/nonsense-on-stilts

<< World is a Story >>


what world do you live

with hawk-eyes on her
safeguarding she
does not create disorder

what world do you create

with yelling in her face
taking away her words
doing so she cannot speak

what world do you hide

convincing her being the barer
of chaos and fear.
then as much:

she will bring the change

—animasuri’22

thinking of Albert Woodfox
passing August 4, 2022
Woodfox, A. (2019). “Solitary. Unbroken by four decades in solitary confinement. My story of transformation and hope.” New York: Grove Press

<< Contextualized Decompositions >>

“ be headed” . digitally photo-edited digital photo . —animasuri’22

PART IIIIIV: prologue

In the world of research and application, there, lies the inflicted “AI.” I, as a flâneuring lay-person, have noticed what seems as polemics, between those experts vehemently promoting neural networks, and they staunchly nuancing any infallibility the latter camp claims, (back to / toward including) symbolic AI.

These authors, researchers, engineers, evangelists and some true believers have, surely unwittingly, poked me into considering a non-AI, digital and uneducated series, slowly collecting my interpretations on this intriguing topic. In simple terms: this is how I have been learning about your field for the past few years. Learning, as a non-machine, non-child, deems not to follow methodological institutional systematic rules, at all times and in all spaces. And then, it also does.

This process, and such output as this one here, is possibly befittingly, or overfittingly, a decomposition on (the) matter; if you will.

This exploration has been going on for a while. At times it was hidden, at times it was an openly trying to be hiding my fear to utter where the (claimed) expert giants roam.

Here, with the seemingly simple entitled photo above and among this text, I nurture a more defined trial. It is intended to be mushroomed over time. This one here, is a poem for you, Giant of Machine and Artifice.

Part IIIIV: perilogue

Where are sets of meaning in the above visualization entitled “ be headed” —- if any meaning at all (to the receiver), while many to its creator and perhaps very different ones to yet other humans or other transforming transcoders?

is it to be found liminally, there in between the words (and the visual) where the artificial space separates “being” from “heading” towards something? be headed is not as be, headed not as be…………headed there.

OR-AND, is it to be found liminally at a dimension unwritten of one word hinted at, torn apart into a conjugation of being and head? Nnnnyes.

Can a network delicately unveil this and other nuanced or simultaneously parallel, hidden, yet to be unveiled or contradictory meanings? Can analysis via formal logic do so and get there?

Can the analysis unveil the unknown unknowns in possible meaning and must these then be accepted as a new ontology to unquestionably submit to?

OR-CONDITIONALLY, under the flag of pragmatic clarity and universal understanding, will an irrational broom be used to batter meaning into convenient consent? Descriptive, imposing, non-negotiably?

There is beauty in plasticity (ANDAND also as a process hinting at ambiguity + yes, don’t be afraid, at some texture of non-transparency) of meaning ANDAND metaphorical neurons ANDAND pragmatic Pierce.

This perhaps to the surprise of the initiated: beauty is sensed even by some of the uninitiated who are not (yet) seeing the enlightening covenants, enabling one seeing beauty whispered in Mathematics, while taming pedestrian and bland math.

Yes, I still lack enlightenment. Mind you, Enlightened One, so do the majority of your fellow humans. What does it then say of your dataset, if your outliers outweigh your desired sample?

The opposite, where one ridicules the other for not seeing one’s aesthetic, might, in its act of debasement by pretending to behead the other, contradict “solving” complexity. Constraining awe is then perceivable as anti-awe.

Is this our collectively carried child’s play at the highest order of human intellect; well-beyond the sphere where I and many more are to be headed? One might wish to circumvent it as such. Though, simultaneously, it might be less intelligent, yet wiser, to remember that debasement is likely the expression sprouted from unrecognized ignorance, imposed on the willingly disagreeable other.

It’s somehow thought so much easier to deny an other any consideration; deny meaning; even among they who unlocked beauty in Mathematics. Where is one’s enlightened insight then? (there, I intuit, lies a delightful paradox).

ANDYET, meaning keeps festering as long as consciousness blooms its spaces, well beyond the visualized linearity or sanctioned connectivity of a (written) syntax, (hierarchical) grammar, semantic (net), and (formalized) logic. Such as any other meaning by any other name is quickly binned, calibrated and celebrated as nonsensical.

Part IIIV: paralogue

Penrose and Hameroff hinted at a transitional in-between. A space where quantum physics and Newtonian physics are “transcoded” (for lack of any hint of substantiated understanding on my part; I am confidently lacking yet open to learning).

Is this what they call a microtubular space; is it a non-computational space? Or so my feeble mind wants to find one simplified meaning, among many more as if vectors upon vectors: pulling, pushing, stretching and contracting. How does pattern AND-OR meaning sprout there; perhaps metaphorically, as a mushroom, screaming relations in subterranean spaces.

If non-binary quantum computing and complexes of computation were to ever be-come com-bined, will logic or metaphorical representations of neural networks be able to be AND not be? Will they then be headed where all possible meaning lays to be captivated, as low hanging fruits, as if possibly decapitating any outliers be-yond reach, and which do not fit their fruity model?

Part IIV metalogue

Meaning is re-imagined, decomposed as a withering mycelial fruit of unknown origin. Beneath the fruits, the networking of “meaning” crosses species (“meaning” is what I anthropomorphically attach to it).

It is a truism, which is possibly hiding further depth, that the signaling occurs across and via the networks themselves. The transitioning of information signals occur in between, and perhaps because of, the negative spaces which tautologically lie outside the recognized held space, and which the physical attributes of the network occupy.

I imagine (and only imagine) the previous as if where space is itself explicitly an informational and meaning-giving, metaphysical, intentional non-architecture. I continue to imagine that this non-architecture is evolved via subtractive and additive synthesis over space and time.

I go deeper down the rabbit hole of my imagination and indulgently give self-satisfying meaning: this non-architecture is imagined as if a medium between quantum physics and the ever so slightly more tangible world.

PART IV: epilogue

Do I *know* and *understand* what I am writing about here?

Answer: no.

I do know and understand that I do not know nor understand. And yet, writing is learning as a snapshot in a process of becoming, if the reader is willing to be informed (or rather: willing to value assigning meaning) as such. This might be what still distinguishes the machine from the human; it does not know that it does not know nor that it does not understand.

Reverting this state back to humans, some who are not knowing they do not know that this could be given meaning to. Meaning as being imitating, inconsiderate, flippant, with pretense, pretentious, delusional, arrogant, having hubris or plainly being (un)intentionally dangerous.

Yes, a human can want to not know what they don’t know. Machines can neither offer this type of deceit. A machine cannot not want to register an input.

As much as their respective opposites, I also imagine this not-knowing and this not-understanding are relational, contextual and adaptive. I find in these meaning by relating back to myself, via self-reflection (however flawed), and (the unwitting) others.

The machine, as a human derivative, is at this stage neither able to express such a verbalization of imaginative meaning-making processes. It is derivative cleverness and hence incomplete and not nuanced to sensibly represent the fullest in-between spaces of human potential meaning-making.

In analogy, we humans are derivatives of the stars, calling a human a star does not make it so ANDYETNOR make it so any less.


PART V: pentalogue

Will the artificial net or the artificial logic, each as a model of the universal “rules” (though what with rules for the non-computational?), then be enabled to identify the (imagined) ability to be, ANDNOT be, contained in one place only, or would we rather loose our heads over this?

while the mathematized DALL-E mangles meaning and defaces human heads into seas of blurred humanization, we humans are sanctioned for playful or surreal or other (un)meaning-making, or exploration thereof at other more or less (costly) dead-ends, leaving serendipitous futures beheaded of meaning to be-come. It does not have to be if we keep our and more so others’ heads on.


Intentionally blurring and poetically yours,

—animasuri’22

post scriptum: I decided not to reference any text implied with (and in between) the above visuals and words.