Tag Archives: aiethics

<< Quantified Efficiency >>

Vlad sat down on that chair, with a sigh,
in what otherwise appeared to him
as eyeing a sightless single-person sofa

“So, I just asked myself,”
he cracked the silence,
“how many hours do I

approximately speak
every work week? Well,
I roughly calculated that

including two cats, two drivers,
two children, one life ‘s partner,
the sporadic delivery person,

a colleague here or there
and my own inner voice
that I speak

about 90 hours a work week.
While that might seem to justify
people stating that I speak too much,

and yet if I cut out my inner voice
it’s only about 6.5 hours a work week
of which about two hours

is spent on the cats,
resulting in about 4.5 hours worth
of human chatter other than with myself.

Of that time I speak
a little less than an hour
with my children per work week.

3.5 hours remain to be assigned.
I calculate a little less than an hour
with morning and evening drivers

and delivery people, in the evening, at times
on the phone with autonomous voices:
should I calculate talking to a phone?

That leaves 2.5 hours.
I speak a little less than an hour
a work week with my colleagues

of whom most I do not speak with
except then the daily warm hello and goodbye.
While with some not even that occurs.

1.5 hours endures the count
of which I spend listening, and replying in mind,
to you for a little less than an hour.

That means I speak about 30 minutes
with you per work week.
That’s around six minutes per work day.

Let’s make it count!” Time being up,
Vlad retreated into silence
with the exception of erecting a sigh

enabling a walk out of the room
voiced by squeaking shoes
and aging joints

                      —animasuri’24

<< Flowers and Rocks >>

beaten drums
beaten competition
beaten child
out of future

beaten models
beaten builds
beaten students
down the street

beaten brainpower
beaten bees
beaten path
swept aside

“and be a simple
kinda man
be somethin’
you love n understand”

b’tween your beats
simply b’tween d’ drops
learn ‘bout somethin’ you
you don’t understand

b’tween your beats
simply b’tween d’ drops
repeat somethin’ you
you don’t understand

        —animasuri’24

<< Type Writer Fauna >>

Put a snake at a type writer
You’ll eventually read something
that stings

Put a horse at a type writer
You wait and see from
canter and gallops winnings appear

Put a monkey at a type writer
You suddenly have to justify
Shakespeare

Put yourself at a type writer
you become a silence monkey-horse
slithering well-formed minotaurs

among words with sweat, flow
as pearls for noses toil
against the forest’s floor

—animasuri’24

<< Humanely Contextualizing ‘AI’ >>

set: 000x

If ‘Odinshühnchen’ were a set of tokens it could be probabilistically reasonable to rehash it to ‘Odin’s hühnchen’  or could functionally be confabulated to become ‘Odin’s chick,’ possibly “hallucinated” to be of chocolate make. It seems when temperatures (t) are set too high (well above 2) a ‘chocolate chicken of Odin’ suddenly seems fair play.

Technology infused poetic thoughts —(by humans for humans) as a discussion and negotiation of nuances of linguistic intuitions and the weighing of words within a compound concept— are dimensions and vectors of what makes us human, well beyond the state of mathematical, engineering and scientific reductions we have access to for the moment. As with the statement “oxygen isn’t enough for humane life,” one without it would be equally challenging as thinking without reductions. And yet, that does neither exclude going beyond either.

Let me elaborate: non-sense, or nonsense, makes us human. Think of a snail intending nonsense; or of a lion. Perhaps a lion cub, or a dolphin might seem to have their moments and yet one might suspect the transposition of human features or human needs for sense-making onto non-human inanimate or animate others. Nonetheless, one might also wonder whether it is humans as a species who take non-sense to very diverse depths or heights. If so, one might assume this is since nonsense can be defined as acts (of which thoughts are a subset) lacking sense, for sense must be there to outlie nonsense away from its median, mean or mode’s stances.

Let us play with sense and nonsense via a loose interpretation of a neurolinguistic concept of “concept blending” while contextualizing the hypes, utopias and dystopias veining the public narrated peripheries of the field of AI. (That’s where this ignorant author roams).

For instance, ‘algorithmic’ we can agree is an adjective.  It intuitively asks for another word to follow. Likely a noun. This sensation is very much as with playing musical chords or sequences that create an anticipation in the listener for some specific chord or note to follow. As it is with bird song. And so it follows…

In this case, let us imagine ‘algorithmic’ is followed by ‘fairness.’ It follows that this would result in the concept (a compound of two words) ‘algorithmic fairness’ just as, for instance, if ‘Odin’ were followed by ‘chicken’ would most likely result in ‘Odin’s chicken.’

Secondly, let us now assume that ‘algorithmic’ can be followed by a number of nouns. This creates a set of concepts that start with the adjective ‘algorithmic.’ For instance, another one is ‘algorithmic governance.’ Sure, one could play a surreal game and suggest ‘algorithmic potato’ or ‘algorithmic chocolate chicken.’

Here thus an intuition is introduced where the weight or the probability of one is higher than the other. And, where the “closeness” of one is closer than the other. ‘algorithmic’ is closer to ‘fairness’ than it is to ‘potato’ or ‘chicken’ and perhaps, to some, less so ‘chocolate.’

This is partly due to one (e.g., ‘fairness’) having been used by other humans (perceived as leading voices or as voices echoing leading voices) while the other is not or probably is not (i.e., ‘chicken’). 

In the ‘algorithmic’ set of concepts with the adjective ‘algorithmic,’ we (too) often decide not to include concepts such as ‘algorithmic potato.’ And yet , the set could still have a number of concepts within it.

 One could decide to notate these concepts differently than how they are traditionally read. So instead of populating this set with constructs such as ‘algorithmic fairness,’ they would be populated with constructs such as ‘fairness, algorithmic_’ and not with ‘potato, algorithmic_’

Though some concepts give the intuition that this notation would not be entirely proper. (Yes, yes, really) Possibly because the link between the word ‘algorithmic’ and the following word or words is stronger or closer. 

For instance: ‘Algorithmic Decision-Making Systems (ADMS)’ It feels as if this concept is less open to be notated in the structural style of ’fairness, algorithmic_’ It might be due to the fact that there is an acronym ‘ADMS’ and its capital letters or it might be something else. It might be because the concept is not only two words and rather three.

A similar intuition occurs with ‘algorithmic outrage deficit.’ And yet a different intuitive oddity occurs with ‘potato, algorithmic_’ let alone ‘chicken, algorithmic_’ and even more so ‘chick, algorithmic_’ not to mention ‘chocolate chick, algorithmic_.’ Earth shattering stuff this is. 

These intuitions introduce inconsistencies into the set. The set which could contain ‘fairness, algorithmic_’ and ‘governance, algorithmic_’ and ‘harm, algorithmic_’ and ‘manipulation, algorithmic_’ and ‘oppression, algorithmic_’ and ‘persuasion, algorithmic_’ and ‘profiling, algorithmic_’ and ‘suggestion, algorithmic_’ and ‘transparency, algorithmic_’ and so on.  

Therein lie, to this ignorant mind, the juices of a type of “concept blending” and perhaps, who knows, of fountains of discomfort and perhaps even anger in they who oppose such human-made blends. Who knows, who’s to chirp about it?

The above might have been conceptually blended (behind the scenes) with a list of bird names translated from German into diverse sets of language, triggering imaginations of sense-making in human sense-making reshuffling, and that in manners that seem to defy sensibility and functions or matrices of probabilistic processes.

‘Transparency, algorithmic_’ is thus questionable here.

—animasuri’24

<< Artificial Sensualities >>

if the friendly colors
and beautiful things as

a prompt injection risk
are concerned with models

for data in the nude
dating toxic hallucinations

thus lies spread before me
spanning space:

handing visualization
and lying with statistics

there is physical harm to be had
when revealing confidential points

on the data corpus
and its improper usage

if junkmail were colored
by the numbers and decaying paper

self-publishing at the input
of recycled sensuality

of a naked typo
of style collage flippantly

inject and sort a snort
it is all the hype again

—-animasuri’24



triggers

Levy, D. A. (1969). the Buddhist Third Class Junkmail Oracle. August 1969. Online: Ubuweb.
https://ubu.com/media/text/vp/buddhist_third_class_junkmail_oracle_aug_1969.pdf

IBM. (2024-03-27). Prompt injection risk for AI. IN: IBM Documentation. IBM Cloud Pak for Data 4.8.x. AI risk atlas. https://www.ibm.com/docs/en/cloud-paks/cp-data/4.8.x?topic=atlas-prompt-injection

<< Men Conducting Artifice >>

Reality as an illusion
—when disembodied—
turns into a submission
to artifices as constructs
of illusion upon illusion upon soothing

“you will find me if you want me
in the garden unless…”

where the simulacra are formed
by those with control over their weights
then truth becomes a shaking of fists
but a setting seeding intention against fate
she loves me, she doesn’t, she loves me, she

“what the hell is he building in there…
he’s hiding from all the rest of us….”

are disconnects of intelligences
undoing interaction of flesh and senses
presenting enslavement as mesmerization
of feedback loops and reductions
to the digital reshuffling agents reassigning values

“happiness is…” poco allegro elongated
can liveliness, yes, in minor fifth’ed be canned

fire, wheel, electricity, one artifice for
kings thrown to dust by creative destruction
and yet dust holds one consistency:
absence of water with water stilling fire,
crackling electricity, conducting artifice

“no, it’s a body of water…
the monuments have been brought down to earth….”

as sediments and sounds of water
from the old pond swooshing,
water creaking, rushing, soothing,
water, whirling, sloshing, water drowning
becoming the overshot water wheel

“Se necesita mucha fortaleza
para levantar….”

bubbles pop as balances between
observations and constructions
distractions away from confident doubt
readily available to anyone
into the puppetry of the happy few

be they chords, be they hursts
Clickedeecklick dripping upwards

onto your sandy banks: immer mid stream

                          —animasuri’24

—-•
some triggers

Abou-Khalil, Rabih. (1992). Sahara. IN: Blue Camel.

van Beethoven, L., Pollini, M. ( ). 33 Variations in C Major, Opus 120 on a waltz by Diabelli: Variation II (Poco Allegro)

Einstürzende Neubauten. (1996). The Garden. IN: Ende Neu.

Garcia-fons, Renaud. (2010). Fortaleza. IN: Méditerranées

Fitzgerald, E., Pass, J. (1976). Nature Boy. IN: Fitzgerald and Pass…Again

Lhamo, Yungchen. (1998). Happiness is… IN: Coming Home.

Mingus, Charles. (1963). Myself when I’m Real. IN: Mingus Plays Piano.

Namchylak, Sainkho. (2016). So Strange! So Strange! IN: like a Bird or Spirit, not a Face.

Vito, Acconci. (2001). The Bristol Project.

Waits, Tom. (1999). What’s he Building? IN: Mule Variations.

<< The Bug of War's Cold Called Love >>

Yesterday I pulled a hair of yours
from the GPU no longer processing
what it was for: the hair,
a single visual of virility, once golden lock

I desire for accuracy, and proper process
I demand cleanliness and method
I own tools in proper Goldilocks zones
yesterday I pulled a hair of yours

it prompted a visual of how I see you
output hidden from alien invasion, how I saw you
within the cranial privacy of my neuro-being
a crack, a rift, the hemispheres appeared

the polis that is my mind partisan’ed
now multidimensional making of truth
conflicting as love-making that nightly chess game
yesterday I pulled a hair of yours from the mainframe

Yesterday I pulled you from the motherboard
one strand one hand one rip one slip
calculations halted, output fuzzed logic
principles and agendas shivered

positivist poles discovered
a third repulsion choreographing
a fourth attraction directing
multitudinous memory of a plucked string resonatingly yours

Yesterday was the day
you were no more, that many years ago, and yet
and yet and yet: here you are my dear:
here you show.

                        —animasuri’24

<< World Model >>

Jacobus was looking for a friend.
Someone who could understand.
Someone, or anything really,
who would share their own experience
Someone Jacobus could relate to.

For years now Jacobus has been living.
Living with hiccups.
It’s interruptive
and jolting joy.
It is a serious matter.

At times Jacobus’ fist
is planted onto the table’s surface.
Cracks occurred,
and that while having a background in engineering
as well as having a degree in medicine:

with a specific focus
on computer science
and then these latest learning technologies
of machines,
anthropomorphically deep

—deeper then the mucosal surfaces of Jacobus’ larynx, glottis and all
—with a specialization in diaphragm irritations and spasms.
with a confidence of authority and funds on call
with pitched, intense unquestioned interest and hypes and
with thrall

Incisively, Jacobus set out to invent a model
one that could generate output
interrupted by hiccups.
Incessantly, incessantly, multimodally: hiccups;
while hurling all water from the soil.

“Drink a cup of water,” said Jacobus,
“take it in.” Knowing full well that didn’t work
for him to date, not at all.
None of that brought him any closer
to that one, idealized breath of peace.

—animasuri’24

—-•
a trigger

Fuchs, T. (2024). Understanding Sophia? On human interaction with artificial agents. IN: Phenom Cogn Sci 23, 21–42 (2024). https://doi.org/10.1007/s11097-022-09848-0

Howick, J., Morley, J., Floridi, L. (2021). An Empathy Imitation Game: Empathy Turing Test for Care- and Chat-Bots. IN: Minds and Machines 31 (3): 457–61. https://doi.org/10.1007/s11023-021-09555-w.

McKeown, G. (2015). Turing’s menagerie: Talking lions, virtual bats, electric sheep and analogical peacocks: Common ground and common interest are necessary components of engagement. IN: International Conference on Affective Computing and Intelligent Interaction (ACII), Xi’an, China, 2015, pp. 950-955, doi: 10.1109/ACII.2015.7344689.

Selinger, E., Dreyfus, H., & Collins, H. (2007). Interactional expertise and embodiment. Studies in History and Philosophy of Science Part A. https://doi.org/10.1016/j.shpsa.2007.09.008